Please have a great command of English with great grammar.

Consider a business with which you are familiar, and which has at least one known, significant competitor. Write a paper that includes the following sections, organized using APA headings (not the Part letter).

Part A: Introduce the paper with the background and information about the business, and the thesis for your paper (1 paragraph).

Part B: Explain, using Teece’s (2010) research article as a basis for your assignment:

1. The business model.
2. The business strategy by:

   Segmenting the market.
   Creating a value proposition for each segment.
   Describing the apparatus to deliver the value.
   Creating the preventative methods to avoid being imitated (p. 180).

3. The business’s competitive advantage, using Teece’s definition and explanation as support (3–5 paragraphs).

Part C: Describe the main advantage your business’s main competitor has against the business.

1. Analyze, using the Red Queen effect articles, [Derfus, Maggitti, Grimm, and Smith (2008) and Giachetti, Lampel, and La Pira (2017)], your business’s potential problem caused by its competition’s solution to the Red Queen effect.
2. Include in your analysis one paragraph that synthetically presents each of the Red Queen articles’ main themes, research findings, and implications. Explain how the concepts of Red Queen effect have grown from 2008 to 2017 (at least 2 paragraphs).
3. In one paragraph, explain your one recommended action as a solution to the problem caused by the competitor’s action. Describe whether or how it could increase competitiveness.

Part D: Conclusion. Summarize the main points of your paper and leave the reader with a thought to go forward with as an implication or recommendation from your ideas and analysis.

Part E:

References

.

• Include a list of all references used.
• Use at least 4 references: the three required readings, and one supporting your information about your selected business.
• Use APA 6th edition on your paper and submit a paper free from errors and of high academic quality.

Refer to the scoring guide to ensure you have covered all the requirements.

References

Derfus, P. J., Maggitti, P. G., Grimm, C. M., & Smith, K. G. (2008). The red queen effect: Competitive actions and firm performance. Academy of Management Journal, 51(1), 61–80.

Giachetti, C., Lampel, J., & Li Pira, S. (2017). Red queen competitive imitation in the U.K. mobile phone industry. Academy of Management Journal, 60(5), 1882–1914.

Teece, D. J. (2010). Business models, business strategy and innovation. Long Range Planning, 43(2–3), 172–194.

Long Range Planning 43 (2010) 172e194 http://www.elsevier.com/locate/lrp

Business Models, Business
Strategy and Innovation

David J. Teece

Whenever a business enterprise is established, it either explicitly or implicitly employs
a particular business model that describes the design or architecture of the value creation,
delivery, and capture mechanisms it employs. The essence of a business model is in de-
fining the manner by which the enterprise delivers value to customers, entices customers
to pay for value, and converts those payments to profit. It thus reflects management’s
hypothesis about what customers want, how they want it, and how the enterprise can
organize to best meet those needs, get paid for doing so, and make a profit. The purpose
of this article is to understand the significance of business models and explore their
connections with business strategy, innovation management, and economic theory.
� 2009 Published by Elsevier Ltd.

Introduction
Developments in the global economy have changed the traditional balance between customer and
supplier. New communications and computing technology, and the establishment of reasonably
open global trading regimes, mean that customers have more choices, variegated customer needs
can find expression, and supply alternatives are more transparent. Businesses therefore need to
be more customer-centric, especially since technology has evolved to allow the lower cost provision
of information and customer solutions. These developments in turn require businesses to re-eval-
uate the value propositions they present to customers e in many sectors, the supply side driven
logic of the industrial era has become no longer viable.

This new environment has also amplified the need to consider not only how to address customer
needs more astutely, but also how to capture value from providing new products and services.
Without a well-developed business model, innovators will fail to either deliver e or to capture e
value from their innovations. This is particularly true of Internet companies, where the creation of
revenue streams is often most perplexing because of customer expectations that basic services
should be free.

0024-6301/$ – see front matter � 2009 Published by Elsevier Ltd.
doi:10.1016/j.lrp.2009.07.003

http://www.elsevier.com/locate/lrp

A business model articulates the logic and provides data and other evidence that demonstrates
how a business creates and delivers value to customers. It also outlines the architecture of revenues,
costs, and profits associated with the business enterprise delivering that value. The different ele-
ments that need to be determined in business model design are listed in Figure 1.

The issues related to good business model design are all interrelated, and lie at the core of the
fundamental question asked by business strategists e how does one build a sustainable competitive
advantage and turn a super normal profit? In short, a business model defines how the enterprise
creates and delivers value to customers, and then converts payments received to profits.1 To profit
from innovation, business pioneers need to excel not only at product innovation but also at busi-
ness model design, understanding business design options as well as customer needs and techno-
logical trajectories. Developing a successful business model is insufficient to assure competitive
advantage as imitation is often easy: a differentiated (and hard to imitate) e yet effective and effi-
cient e business model is more likely to yield profits. Business model innovation can itself be a path-
way to competitive advantage if the model is sufficiently differentiated and hard to replicate for
incumbents and new entrants alike.

In essence, a business model [is] a conceptual, rather than financial,

model of a business.

In essence, a business model embodies nothing less than the organizational and financial ‘archi-
tecture’ of a business.2 It is not a spread sheet or computer model, although a business model might
well become embedded in a business plan and in income statements and cash flow projections. But,
clearly, the notion refers in the first instance to a conceptual, rather than a financial, model of a busi-
ness. It makes implicit assumptions about customers, the behavior of revenues and costs, the

Figure 1. Elements of business model design

Long Range Planning, vol 43 2010 173

changing nature of user needs, and likely competitor responses. It outlines the business logic re-
quired to earn a profit (if one is available to be earned) and, once adopted, defines the way the en-
terprise ‘goes to market’. But it is not quite the same as a strategy: the distinction and the
relationship between the two will be discussed later.

Despite lineage going back to when societies began engaging in barter exchange, business models
have only been explicitly catapulted into public consciousness during the last decade or so. Driving
factors include the emerging knowledge economy, the growth of the Internet and e-commerce, the
outsourcing and offshoring of many business activities, and the restructuring of the financial ser-
vices industry around the world. In particular, the way in which companies make money nowadays
is different from the industrial era, where scale was so important and the capturing value thesis was
relatively simple i.e. the enterprise simply packed its technology and intellectual property into
a product which it sold, either as a discreet item or as a bundled package. The existence of electronic
computers that allow low cost financial statement modeling has facilitated the exploration of alter-
native assumptions about revenues and costs.

Additional impetus has come from the growth of the Internet, which has raised anew, and in
a transparent way, fundamental questions about how businesses deliver value to the customer,
and how they can capture value from delivering new information services that users often expect
to receive without charge. It has allowed individuals and businesses easy access to vast amounts
of data and information, and customer power has increased as comparison shopping has been
made easier. In some industries, such as the recording industry, Internet enabled digital downloads
compete with established channels (such as physical product sales) and, partly because of the ubiq-
uity of illegal digital downloading, the music recording industry is being challenged to completely
re-think its business models. The Internet is not just a source of easy access to digital data; it is also
a new channel of distribution and for piracy which clearly makes capturing value from Internet
transactions and flows difficult for recording companies, performers and songwriters alike. More
generally, the Internet is causing many ‘bricks and mortar’ companies to rethink their distribution
strategies e if not their whole business models.

Notwithstanding how the Internet has devastated the business models of industries like music re-
cording and news, internet companies themselves have struggled to create viable business models. In-
deed, during the dot.com boom and bust of 1998e2001, many new companies with zero or negative
profits (and unprecedentedly low revenues) sought financial capital from the public markets, which e
at least for a short while e accommodated them. Promoters managed to persuade investors that tra-
ditional revenue and profitability models no longer applied e and that the dot.com companies would
(eventually) figure out (highly) profitable business models. Few have, causing one commentator to
remark that ‘the demise of a popular but unsustainable business model now seems inevitable’.3

No matter what the sector, there are criteria that enable one to determine whether or not one has
designed a good business model. A good business model yields value propositions that are compel-
ling to customers, achieves advantageous cost and risk structures, and enables significant value cap-
ture by the business that generates and delivers products and services. ‘Designing’ a business
correctly, and figuring out, then implementing e and then refining e commercially viable archi-
tectures for revenues and for costs are critical to enterprise success. It is essential when the enter-
prise is first created; but keeping the model viable is also likely to be a continuing task. Superior
technology and products, excellent people, and good governance and leadership are unlikely to pro-
duce sustainable profitability if business model configuration is not properly adapted to the com-
petitive environment. Some preliminary criteria for business model design are suggested
throughout this article, and summarised in a later section.

The concept of a business model has no established theoretical

grounding in economics or in business studies.

174

Business Models, Business Strategy and Innovation

Business models e the theoretical foundation
The concept of a business model lacks theoretical grounding in economics or in business studies.
Quite simply there is no established place in economic theory for business models; and there is not
a single scientific paper in the mainstream economics journals that analyses or discusses business
models in the sense they are defined here. (Possible exceptions are the literature on investment
in basic research, which economists recognize as being unsupported by private business models
(see below), and the literature on bundling, inasmuch as it deals e indirectly e with different rev-
enue models.) The absence of consideration of business models in economic theory probably stems
from the ubiquity of theoretical constructs that have markets solving the problems that e in the real
world e business models are created to solve.

Economic theory implicitly assumes that trades take place around tangible products: intangibles
are, at best, an afterthought. In standard approaches to competitive markets, the problem of cap-
turing value is quite simply assumed away: inventions are often assumed to create value naturally
and, enjoying protection of iron-clad patents, firms can capture value by simply selling output in
established markets, which are assumed to exist for all products and inventions. Thus there are no
puzzles about how to design a business e it is simply assumed that if value is delivered, customers
will always pay for it. Putting so called ‘public goods’ and ‘free rider’ issues to one side, business
models are quite simply redundant because producers/suppliers can create and capture value simply
through disposing their output at competitive market prices. Such models clearly assume away the
essential business design issues that are the subject of this article.

In short, figuring out business models for a new or existing product or business is an unnecessary
step in textbook economics, where it is not uncommon to work with theoretical constructs which
assume fully developed spot and forward markets, strong property rights, the costless transfer of
information, perfect arbitrage, and no innovation.4 In mainstream approaches, there is simply
no need to worry about the value proposition to the customer, or the architecture of revenues
and costs, or about mechanisms to capture value.5 Customers will buy if the price is less that
the utility yielded; producers will supply if price is at or above all costs including a return to
capital e the price system resolves everything and business design issues simply don’t arise.

But general equilibrium models, with (one-sided) markets and perfect competition are a carica-
ture of the real world. Intangible products are in fact ubiquitous, two-sided markets are common,
and customers don’t just want products; they want solutions to

their perceived needs.

In some
cases, markets may not even exist, so entrepreneurs may have to build organizations in order to
perform activities for which markets are not yet ready. Accordingly, in the real world, entrepreneurs
and managers must give close consideration to the design of business models and even to building
businesses to execute transactions which cannot yet be performed in the market.

Equilibrium and perfect competition are a caricature of the real

world. customers don’t just want products; they want solutions to

their perceived needs.

It’s also true that business models have no place within the theoretical constructs of planned
economies (just as in a perfectly competitive economy). While central planners do need to under-
stand the stages in the production system, in a supply driven system e where consumers merely get
what the system produces e business models simply aren’t necessary. There is no problem associ-
ated with producers capturing value because value doesn’t even have to be captured; the state de-
cides what and how to produce, and how to pay for it all.

While business models have no place in economic theory, they likewise lack an acceptable place
in organizational and strategic studies, and in marketing science. However, there has been some
limited discussion and research on new organizational forms. Williamson, for instance, recognizes

Long Range Planning, vol 43 2010 175

that ‘the 1840s marked the beginning of a great wave of organizational change that has brought us the
modern corporation’.6 As discussed earlier, new organizational forms can be a component of a busi-
ness model;7 but organizational forms are not business models. Clearly, the study of business
models is an interdisciplinary topic which has been neglected e despite their obvious importance,
it lacks an intellectual home in the social sciences or business studies. This article aims to help rem-
edy this deficiency.

Examples of business models
Business models are necessary features of market economies where there is consumer choice, trans-
action costs, and heterogeneity amongst consumers and producers, and competition. Profit seeking
firms in competitive environments will endeavor to meet variegated consumer wants through the
constant invention and presentation to the consumer of new value propositions. Business models
are often necessitated by technological innovation which creates both the need to bring discoveries
to market and the opportunity to satisfy unrequited customer needs. At the same time, as indicated
earlier, new business models can themselves represent a form of innovation. There are a plethora of
business model possibilities: some will be much better adapted to customer needs and business en-
vironments than others. Selecting, adjusting and/or improving business models is a complex art.
Good designs are likely to be highly situational, and the design process is likely to involve iterative
processes. New business models can both facilitate and represent innovation e as history
demonstrates.

Traditional industries
A striking early American example of 19th century business model innovation was Swift and Com-
pany’s ‘reengineering’ of the meat packing industry. Prior to the 1870s, cattle were shipped live by
rail from the Midwestern stockyard centers like Omaha, Kansas City and Chicago to East Coast
markets where the animals were slaughtered and the meat sold by local butchers. Gustavus Swift
sensed that if the cattle could be slaughtered in the Midwest and shipped already dressed to distant
markets in refrigerated freight cars, great economies in ‘production’/centralization and transporta-
tion could be achieved, along with an improvement in the quality of the final product.

Swift’s new business model quickly displaced business models involving a network of shippers,
East Coast butchers and the railroads. His biggest challenge was the absence of refrigerated ware-
houses to store the beef near point of sale, which were not part of the existing distribution system.
Swift set about creating a nationwide web of refrigerated facilities, often in partnerships with local
jobbers. ‘Once Swift overcame the initial consumer resistance to meat slaughtered days before in distant
places, his products found a booming market because they were as good as freshly butchered meats and
were substantially cheaper e Swift’s success quickly attracted imitators e By the 1890s, men like Phillip
Armour had followed on Swift’s heels’.8

A more recent example is containerization. Malcolm McLean, owner of a large U.S. trucking
company, was convinced that conventional shipping was highly inefficient because shipping com-
panies typically broke bulk at dockside, and cargo ships spent most of their time in port being
loaded or unloaded. In 1955 he hired an engineer to design a road trailer body that could be
detached from its chassis and stacked on ships. McLean acquired a small steamship company,
renamed it Sea-Land Industries (it eventually became absorbed into the Maersk Line). He devel-
oped steel frames to hold the containers, first on the top decks of tankers, and then on the world’s
first specialized cellular containership, the Gateway City, launched in 1957. To promote the stan-
dardization necessary to develop the industry, McLean made Sea-Land’s patents available royalty
free to the International Standards Organization (ISO). Sea-Land began service on North Atlantic
routes in 1966. When R. J. Reynolds bought Sea-Land for $530 million in 1969, McLean received
$160M for his share and retired.9

Another U. S. example of successful business model innovation is Southwest Airlines, where the
founder surmised that most customers wanted direct flights, low costs, reliability and good

176 Business Models, Business Strategy and Innovation

customer service, but didn’t need ‘frills’. To achieve these goals, Southwest eschews the hub-and-
spoke model associated with alliances, nor does it allow interlining of passengers and baggage,
or sell tickets through travel agencies e all sales are direct. Aircraft are standardized on the Boeing
737, allowing greater efficiency and operating flexibility. Southwest’s business model e which was
quite distinct from those of the major carriers e followed elements of a discount airline model first
pioneered in the U.K. by Freddie Laker. Although Laker Airways eventually failed e as did other
early followers in the U.S. such as People’s Express e Easy Jet has implemented a similar model
in Europe, so far successfully.

The ‘razor-razor blade model’ is another classic (and quite generic) case of a well known business
revenue model (which is just one component of a business model), which involves pricing razors
inexpensively, but aggressively marking-up the consumables (razor blades). Jet engines for com-
mercial aircraft are priced the same way e manufacturers know that engines are long lived, and
maintenance and parts is where Rolls Royce, GE, Pratt & Whitney and others make their money.
So engines are sold relatively inexpensively e but parts (and service) involve considerable mark-ups
and represent an income stream that may continue for decades.

The ’razor-razor blade model’ is a classic business revenue model … jet

engines for commercial aircraft are priced the same way.

In the sports apparel business, sponsorship is a key component of today’s business models. Nike,
Adidas, Reebok, Canterbury, and others sponsor football and rugby clubs and teams, providing kit
and sponsorship dollars as well as royalties streams from the sale of replica products. After building
brand on the field, these companies endeavor to leverage their brand into off-field products, often
with considerable success. On-field sponsorship is almost a sine qua non for brand authenticity.
However, this model is readily imitated, and its viability for any particular apparel company
depends on the sponsor’s particular abilities to leverage on-field sponsorships into off-field sales.
Relationships with clubs, teams, and with team managers and club owners become important in
the mix.

Performing artists have several business models they can employ. Their revenue sources might
include live productions, movies, sale of physical CDs through stores or online music sales through
virtual stores such Apple’s iTunes.10 Stars might decide to use concerts as their main revenue gen-
erator, or to spend less time performing and more in the recording studio, using concerts primarily
to stimulate sales of recordings. In earlier days when piracy was limited, the Beatles demonstrated
that stars could quit live performances and continue to do well on royalties from the sale of re-
corded music. Then, in the 80s and 90s, the music video became an important source of revenue,
and more recently, ‘soundtracks’ to video games have become a significant source of revenue for
some artists. In short, multiple revenue streams are available, and the particular revenue model em-
ployed can depend on the marketplace, on a star’s contextual talents and preferences, and on the
quality of copyright protection afforded to recorded music.

Business models must morph over time as changing markets, technologies and legal structures
dictate and/or allow. For instance, the business model that U.S. investment banks had employed
for almost 20 years largely disappeared in 2008. From at least the 1950s through the 1990s, the in-
vestment banking function usually generated most of the banks’ revenues. However, for Goldman
Sachs (arguably the industry leader) that figure had fallen to 16% by 2007, while revenues from
trading and principal investment had grown to 68%, leading it and other investment banks to
morph their business models into something quite different e and more risky e than traditional
investment banking. Subprime mortgages and other problematic assets became securitized and in-
jected into the system, encouraged by Freddie and Fannie (and by Congress) with results that sub-
sequently hit the headlines. In September 2008, Goldman Sachs and Morgan Stanley (the last two

Long Range Planning, vol 43 2010 177

independent investment banks left standing in the U.S. after the takeover of Bear Sterns by JP Mor-
gan Chase, the bankruptcy of Lehman Brothers, and Merrill Lynch’s absorption by Bank of Amer-
ica) converted themselves into federally chartered commercial banks. By accepting government
regulation by the FDIC, Goldman Sachs and Morgan Stanley will need to maintain lower leverage,
and accept lower risk and lower returns. In their need for a source of stable funds, both have (albeit
reluctantly) made significant business model changes e in short, they have been obliged to abandon
their old models entirely.

The information/internet industries
As noted earlier, the information industries have always raised challenging business model issues
because information is often difficult to price, and consumers have many ways to obtain certain
types without paying. Figuring out how to earn revenues (i.e. capture value) from the provision
of information to users/customers is a key (but not the only) element of business model design
in the information sector. The rules for strategic engagement promulgated by Shapiro and Varian
are core elements of strategy in the information services sector.11

As traditional information providers, newspapers have employed a revenue model for decades in
which the paper is sold quite inexpensively (usually at a nominal level, insufficient to cover costs),
while publishers looked to advertising revenue to cover remaining costs plus provide a profit. In
recent years, this business model has been undermined by websites like eBay and Craigslist that
have siphoned off advertising revenues from job and real estate listings and classified ads: many
newspapers have gone out of business.

The Internet has enabled traditional industries like DVD rentals to adopt a more modern on-line
posture. Netflix (http://www.netflix.com) enables customers to order DVDs on-line and have expe-
dited delivery by the U.S. Mail as a more convenient alternative to going to a rental facility, renting
the DVD, and returning it several days later. Monthly fees are what sustain Netflix.

The emergence of the Internet, Napster and its clones has obliged music recording companies to
rethink their business models, which they have been doing along several fronts. On one front, they
are moving to greatly increase the royalty rate for Internet ‘broadcast’ of their content, while on
another, they are moving to capture advertising revenues associated with that content. For instance,
MySpace Music (http://music.myspace.com) enables users to listen to songs from Universal, Sony
BMG and Warner Music, and provides free advertising-supported streaming, with easy access to
Amazon.com for music purchases. Another example is the Nokia ‘Comes with Music’ (CWM)
handset that comes with ‘free’, unlimited music downloads for a year, with Nokia passing on
a fee to the recording companies.

A recent example of an Internet business model is Flickr (www.flickr.com), which has been described
as ‘a poster child for Web 2.0 [offering] users a way to share photos easily’.12 Flickr’s friendly and easy-to-use
web interface and its free photo management and storage service are noted as great examples of a Web 2.0
‘freemium’ (free and premium) business model, characterized by Fred Wilson as:

‘Give your service away for free, possibly ad supported but maybe not, acquire a lot of customers
very efficiently through word of mouth, referral networks, organic search marketing, etc., then
offer premium priced value added services or an enhanced version of your service to your
customer base’.

The Flickr business model (which actually evolved from gaming to on-line photo sharing, harness-
ing user feedback generated through blogs) essentially gives away the services that amateur photogra-
phers want most: photo sharing, on-line storage, indexing and tagging. Shuen notes that low cost
on-line distribution and marketing and investment are associated with ‘revenue from multiple streams,
including value-added premium services and customer acquisition.’ Flickr’s multiple revenue stream
business model involves collecting subscription fees, charging advertisers for contextual advertising,
and receiving sponsorship and revenue-sharing fees from partnerships with retail chains and comple-
mentary photo service companies. Yahoo bought Flickr in March 2005 for tens of millions of dollars.

178 Business Models, Business Strategy and Innovation

http://www.netflix.com

http://music.myspace.com

http://Amazon.com

http://www.flickr.com

companies can adopt business models [e.g. Freemium or multiple

revenue stream models] pioneered in one space into another.

A business model pioneered by one company in one space may be adopted by another company
in another space. The ‘freemium’ model has been adopted by Adobe (for its PDF reader), Skype and
MySpace, while Outshouts Inc (www.outshouts.com) has applied Flickr’s multiple revenue streams
model to on-line Web videos, allowing users to personalize and disseminate videos for business or
consumer purposes. While it is common with Internet start-ups, the multiple revenue stream ap-
proach is by no means new. Besides theatrical releases and looking to exploit an obvious extra rev-
enue stream e the sequel e movie studios have long sought revenues from ‘ancillary’ licensing
(toys, T-shirts, lunchboxes, backpacks), and more recently from video games and soundtracks.

Freemium business models are also deployed by a large number of software companies (such as
Linux, Firefox, and Apache) who operate in the open source marketplace. The standard form (or
‘kernel’) of the software is licensed under an open source license and then a premium version with
additional features and/or associated services is made available under commercial license terms.
One theory is that ‘vendors’ get customers (often, and ideally with the IT organization bypassing
Procurement Departments altogether e because, after all, the software is ‘free’) hooked on the
free product, and then subsequently convert them into paying customers through the sale of com-
plementary software and/or service. However, conversion rates to paying customers have been
poor, and it’s not clear the model works.

The discussion so far has focused mainly on the impact of technology on value and its delivery.
However technology can have an equally transformative effect on the cost side of the business model.
New ‘cloud-based’ computing models, for example, remove the need for small companies to invest up-
front in expensive servers e instead they can buy server capacity in small slices, as needed, according to
their monthly needs. The size of such slices continues to shrink e services such as Amazon’s EC2, for
example, even allow customers to buy virtual server capacity for a single transaction, measured in mil-
liseconds. This kind of innovation transforms previous ‘fixed plus variable’ cost models into entirely
variable cost models, greatly improving efficiency and reducing early-stage capital requirements.

Business models, strategy and sustainable competitive advantage
A business model articulates the logic, the data, and other evidence that support a value proposition
for the customer, and a viable structure of revenues and costs for the enterprise delivering that
value. In short, it’s about the benefit the enterprise will deliver to customers, how it will organize
to do so, and how it will capture a portion of the value that it delivers. A good business model will
provide considerable value to the customer and collect (for the developer or implementor of the
business model) a viable portion of this in revenues. But developing a successful business model
(no matter how novel) is insufficient in and of itself to assure competitive advantage. Once imple-
mented, the gross elements of business models are often quite transparent and (in principal) easy to
imitate e indeed, it is usually just a matter of a few years e if not months e before an evidently
successful new business model elicits imitative efforts. In practice, successful business models very
often become, to some degree, ‘shared’ by multiple competitors.

A business model is more generic than a business strategy. Coupling

strategy and business model analysis is needed to protect competitive

advantage resulting from new business model design.

Long Range Planning, vol 43 2010 179

http://www.outshouts.com

As described, a business model is more generic than a business strategy. Coupling strategy anal-
ysis with business model analysis is necessary in order to protect whatever competitive advantage
results from the design and implementation of new business models. Selecting a business strategy
is a more granular exercise than designing a business model. Coupling competitive strategy analysis
to business model design requires segmenting the market, creating a value proposition for each seg-
ment, setting up the apparatus to deliver that value, and then figuring out various ‘isolating mech-
anisms’ that can be used to prevent the business model/strategy from being undermined through
imitation by competitors or disintermediation by customers.13

Strategy analysis is thus an essential step in designing a competitively sustainable business model.
Unless the business model survives the filters which strategy analysis imposes, it is unlikely to be
viable, as many business model features are easily imitated. For instance, leasing vs. owning is an
observable characteristic of business models that competitors can replicate. The ‘newspaper revenue
model’ e i.e. low cost for the newspaper, use of advertising (including classifieds) to help cover the
costs of generating content e is easy to replicate, and has been implemented with little variation in
thousands of geographically separate ‘markets’ throughout the world.

Having a differentiated (and hard-to-imitate) e but at the same time effective and efficient e
architecture for an enterprise’s business model is important to the establishment of competitive ad-
vantage. The various elements need to be cospecialized to each other, and work together well as
a system. Both Dell Inc. and Wal-Mart have demonstrated the value associated with their business
models (while Webvan and many other dotcoms demonstrated just the opposite). Dell and Wal-
Mart’s business models were different, superior, and required supporting processes that were
hard for competitors to replicate (at least in the U.S. e elsewhere, new entrants could adopt key
elements of the model and pre-empt Wal-Mart, as Steven Tindall has demonstrated so ably in
New Zealand with ‘The Warehouse’). Both Dell and Wal-Mart have also constantly adjusted and
improved their processes over time. Michael Dell, founder of Dell, notes:

This belief e that by working directly with customers we could get them technology faster, provide
a better level of service, and provide better value e was the basis of the business e the fundamental
business system was quite powerful and delivered lots of value to our customers e we screwed up lots
of things, but the one thing we got right was this core business model, and it masked any other
mistakes .14

Dell’s competitors were incumbents who had difficulty in replicating its strategy, as selling direct
to customers would upset their existing channel partners and resellers: as a new entrant, Dell had no
such constraints. Another critical element of Dell’s success, beyond the way it organized its value
chain, was the choice of products it sold through its distribution system. Over time, Dell developed
(dynamic) capabilities around deciding which products to build beside desktop and laptop com-
puters, and has since added printers, digital projectors and computer-related electronics. Of course,
the whole strategy depended on the availability of numerous non captive suppliers able to produce
at very competitive prices.

Magretta points out that the business model of discount (big box) retailing had been around long
before Wal-Mart founder Sam Walton (in his words) ‘put good sized stores into little one-horse towns
which everybody else was ignoring’.15 Once in place, the towns Wal-Mart had selected were too small
to support another similar sized store, so a difficult to replicate first mover advantage had been cre-
ated. Wal-Mart promoted national brands at deep discounts, supported by innovative and lean pur-
chasing logistics and IT systems: these were elements of its strategy that made its business model
difficult to imitate.

Search engine development and the Google story is another interesting business model illus-
tration. Early efforts in this field, including Lycos, Excite, Alta Vista, Inktomi and Yahoo, would
find lots of information e perhaps too much e and present it to users in an unhelpful manner,
with maybe thousands of results presented in no discernible or useful order. Alta Vista presented
links, but without using them as aids to searching. Larry Page, one of the founders of Google,

180 Business Models, Business Strategy and Innovation

surmised that counting links to a website was a way of ranking its popularity (much like higher
citation counts in scientific journals point to more important contributions to the literature),
and decided to use the number of links to important sites as a measure of priority. Using this link
based approach, Page and his colleagues at Google devised an Internet site ranking system e the
PageRank algorithm e which went on to be their core product/service offering, and one which
has proved very valuable to users. The challenge was to tune the product offering and devise
a business model to capture value, which was not easy in a world in which consumers expected
search to be free.

The business model developed around Google’s product/service innovation required heavy in-
vestment in computing power as well as in software. Google writes its own software and (remark-
ably) builds its own computers. It takes advantage of its considerable computing power to count
words and links, and to combine information about words and links. This allows the Google search
engine to take more factors into account than others currently in the market. The Google revenue
model eschewed funding from advertisers: directed search biased to favor advertisers was perceived
by Google’s founders as degrading to the integrity of the search process and to its emerging brand.
Accordingly, it decided that the essence of its revenue model would be sponsored links i.e. no pop
ups or other graphics interfering with the search. In short, Page and Brin found a way to accom-
modate advertising (thereby enabling revenue generation) without subtracting from the search ex-
perience, and arguably enhancing it.16 However, they also adopted an integrated approach (by
fulfilling their own software and hardware requirements) to keep control of their product/service
offering, ensuring its delivery and its quality.

Business model choices define the architecture of the business .

expansion paths develop from there on out.

Business model choices define the architecture of the business, and expansion paths develop from
there on out. But once established, enterprises often encounter immense difficulty in changing busi-
ness models e witness the difficulties American Express and Discover Card have experienced in try-
ing to morph to hybrid models where they issue cards themselves while simultaneously looking to
persuade banks as partners to act as card issuers for them. This is clearly incongruous e their main
competitors (Visa and MasterCard, who provide network services only and don’t compete with
banks in issuing credit cards) are not hobbled by such relationship conflicts, and are clearly likely
to be the bank’s preferred partners. Thus American Express and Discover are unlikely to have (and
indeed have not had) much success trying to replicate the Visa/MasterCard business model while
still maintaining their own internal issuing and acquiring functions.17

In short, innovating with business models will not, by itself, build enterprise-level competitive advan-
tage. However, new business models, or refinements to existing ones, like new products themselves, often
result in lower cost or increased value to the consumer; if not easily replicated by competitors, they can
provide an opportunity to generate higher returns to the pioneer, at least until their novel features are
copied. These issues are summarized in Figure 2 and explored in more detail later.

Barriers to imitating business models
This section attempts to distil those factors that affect the ease or otherwise of imitating business
models. At a superficial level all business models might seem easy to imitate e certainly the basic
idea and the business logic behind a new model is unlikely itself to enjoy intellectual property protec-
tion. In particular, a new business model, being more general than a business method, is very unlikely
to qualify for a patent, even if certain business methods underpinning it may be patentable. Descrip-
tions of a business model may enjoy copyright protection, but that is unlikely to be a barrier to

Long Range Planning, vol 43 2010 181

Figure 2. Steps to achieve sustainable business models

copying its basic core ‘idea’. What then is it, if anything, that is likely to impede the copycat behavior
that can so quickly erode the business model pioneer’s advantage? Three factors would seem to be
relevant.

First, implementing a business model may require systems, processes and assets that are hard to
replicate e such was the situation with potential entrants into the towns too small to sustain
a Wall-mart competitor. Similarly, while at some level Dell Computer’s direct-to-user (consumers
and businesses) business model is obvious (you simply disintermediate wholesalers and retailers),
when Gateway Computers tried to implemented a similar model, their failure to achieve anywhere
near Dell’s performance levels has been attributed to the inferior implementation of processes. Ca-
pabilities matter. Likewise, when Netflix pioneered delivery of DVDs by mail using a subscription
system, Blockbuster video responded with a similar offering. But Netflix held on to its lead, both
because it was not handicapped by Blockbuster’s cannibalization concerns, and because it had pat-
ents on the ‘ordered list’ (which it later accused Blockbuster of infringing) by which subscribers
indicated online their movie preferences.

Second, there may be a level of opacity (Rumelt has referred to this opacity as ‘uncertain imita-
bility’) that makes it difficult for outsiders to understand in sufficient detail how a business model is
implemented, or which of its elements in fact constitute the source of customer acceptability.18

Third, even if it is transparently obvious how to replicate a pioneer’s business model, incumbents
in the industry may be reluctant to do so if it involves cannibalizing existing sales and profits or
upsetting other important business relationships. When incumbents are constrained in this way,
the pioneer of a new business model may enjoy a considerable period of limited competitive re-
sponse. Notwithstanding these constraints, competition is likely to be vigorous because other
new entrants, similarly unconstrained by incumbency and cannibalization anxieties, will be equally
free to enter.

Business model learning
The moves made by an incumbent competitor to overcome such barriers to respond to Netflix’s
entry into DVD rentals provide an interesting illustration of Business Model learning and adjust-
ment. To respond to Netflix’s competitive inroads into its DVD store-rental model, Blockbuster
purchased assets from NetLearn in April 2002, including those of DVDRentalCentral.com,

182 Business Models, Business Strategy and Innovation

http://DVDRentalCentral.com

a subscriber-based online DVD rental service, which it renamed FilmCaddy and operated separately
from the rest of the Blockbuster business. In August 2004, Blockbuster shut down FilmCaddy and
launched Blockbuster Online, its new online rental service that allowed customers to rent unlimited
DVDs (three at a time) for a monthly fee. Its initial plan included no due dates or extending view-
ing fees, and also gave subscribers two free in-store movie rentals each month. In November 2006, it
launched Blockbuster Total Access, coupling its online business with its in-store capabilities to al-
low online customers the option of returning their DVDs through the mail or exchanging them for
free-in-store movie rentals at over 5000 Blockbuster stores.

Clearly, most elements of the Netflix business model were relatively easy to copy, and, although
Blockbuster was undoubtedly constrained by the cannibalization of its in-store rentals by its online
business, these moves reflected its attempts to respond (defensively) to Netflix. Netflix had figured
out an approach and made the investments required to establish the online market. But Blockbuster
responded by leveraging its brand equity and its network of physical stores to try to capture value
from a modified version of the model Netflix had created: at minimum, it was intent on minimizing
damage to its in-store franchise. Its guiding principle in responding appeared to be to offer cus-
tomers all the functionality of Netflix plus several distinguishing features e associated with using
its retail store footprint e which Netflix couldn’t easily match. Blockbuster’s stores also comple-
mented its online strategy, by offering customers a choice of how to return their rented DVDs.
While Netflix had no retail presence with which to respond to this element of Blockbuster’s offering
directly, it did have some limited patent protection, with two patents that provided its business
model some protection e in particular, to its ‘ordered list’ for movie selection. While these patents
did not cover online DVD rental per se, they did cover methods allowing users to pay a flat fee to
have a maximum number of movies out at any one time, and to return a fixed number of movies
within a fixed time period.

In short, Blockbuster implemented a close facsimile of the Netflix business model (even its web-
site was very similar, featuring stars, recommendations, box shots and the ‘dynamic queue’) and
achieved reasonable success, undoubtedly blunting Netflix’s growth. While Blockbuster Online
was a good defensive move, Netflix’s pioneering status and its capacity to improve its business
model, and enforce its patents, has helped undergird its competitive advantage.

technological innovation does not guarantee business success e new

product development efforts should be coupled with a business

model defining their ’go to market’ and ’capturing value’ strategies.

Business models to capture value from technological innovation

The profiting from innovation framework
Figuring out how to capture value from innovation is a key element of business model design. This
is a topic on which this author has written extensively, although the treatment hitherto was not
couched in the language of business model design. This section is more forthright in that regard.

Every new product development effort should be coupled with the development of a business
model which defines its ‘go to market’ and ‘capturing value’ strategies. Clearly technological inno-
vation by itself does not automatically guarantee business or economic success e far from it. This
was a theme in the author’s earlier work on ‘Profiting from Innovation’,19 which outlined a contin-
gent approach with respect to how to organize the production system/value chain, taking into ac-
count the ‘appropriability regime’ and the innovator’s prior asset positioning. Notwithstanding that
scholars have recognized that technological innovation without a commercialization strategy is as

Long Range Planning, vol 43 2010 183

likely to lead to the (self-) destruction of creative enterprises as it is to profitable (Schumpeterian)
creative destruction, technological innovation is often assumed by some to lead inexorably to com-
mercial success. It rarely does. When executives think of innovation, they all too often neglect the
proper analysis and development of business models which can translate technical success into
commercial success. Good business model design and implementation, coupled with careful stra-
tegic analysis, are necessary for technological innovation to succeed commercially: otherwise,
even creative companies will flounder. Quintessential examples of firms that succeeded at techno-
logical innovation but failed to get the business model and the technology strategy right included
EMI (the CAT scanner) and Xerox (the personal computer).20

But there are a plethora of other examples too. Eli Whitney’s 1793 invention of the cotton gin greatly
increasing the ease with which cotton could be separated from the pod e but still he died a poor man.
Even Thomas Edison e with his portfolio of 1000+ patents and personal fame from inventing a durable
electric light bulb, electricity as a system, motion pictures and phonographs e failed commercially on
many fronts. For example, he abandoned the recording business after arguably failing to get its
business model right by insisting that Edison disks be designed to work only on Edison phonographs
(although his early phonograph also suffered from poor sound reproduction, recordings that were too
brief, and cylinders that could only survive a few playings). In short, getting the business model and the
technology strategy right is necessary to achieve commercial viability if sustainable competitive advan-
tage is to be built and innovators are to profit from their innovations.

Figuring out how to deliver value to the customer e and to capture value while doing so e are
the key issues in designing a business model: it is not enough to do the first without the second. The
imperfections in the market for knowhow make capturing value from its production and sale in-
herently difficult,21 and may often necessitate a business model where knowhow is bundled into
products and complementary assets used to realize value to the innovator. This involves some of
the trickiest and most frustrating issues that entrepreneurs and managers must address.

The Profiting from Innovation framework is an effort to help entrepreneurs and strategists figure
out appropriate business model/designs and technology strategies by delineating important features
of business model choice, and predicting the outcomes from those choices. The framework employs
contracting theory,22 and recognizes two extreme modes (models) by which innovators can capture
value from innovation:

� At one end of the scale stands the integrated business model, in which an innovating firm bun-
dles innovation and product together, and assumes the responsibility for the entire value chain
from A to Z including design, manufacturing, and distribution. Clearly, companies that have the
right assets already in place are well equipped to do this; but the framework also indicates when
the internal development and commercialization strategy is a necessity.

� The other extreme case is the outsourced (pure licensing) business approach, one that has been
embraced by a number of companies, like Rambus (semiconductor memory) and Dolby (high
fidelity noise reduction technology). With respect to licensing versus internal commercialisation
by the innovator, the framework yields answers calibrated according to the strength of the ap-
propriability/intellectual property regime. Thus one could license e and expect the licensing
model to work e only if one had strong intellectual property rights: without them the licensee
might well be the one who captures value, at the expense of the innovator.

� In between there are hybrid approaches involving a mixture of the two approaches (e.g. out-
source manufacturing; provide company owned sales and support). Hybrid approaches are
the most common, but they also require strong selection and orchestration skills on the part
of management.23

[a] licensing model [will only] work [with] strong intellectual property

rights. [otherwise] the licensee will capture value, not the innovator.

184 Business Models, Business Strategy and Innovation

The Profiting from Innovation framework can thus be considered as a tool to help design busi-
ness models, and using it allows one to map business model selection to type of innovation, while
simultaneously enabling one to figure out where intellectual property monetization through licens-
ing is likely to be viable, and where it’s not, or where some kind of vertical integration is indicated.24

Although, (by construction) it is silent on many issues such as market segmentation and the choice
of product features, it nevertheless can provide insights into how a value chain ought to be
assembled. And it can predict winners and losers from the competitive process in the context where
a customer need is being met.

‘Public’ goods and the bundling and unbundling of inventions and products
Inventors and innovators rarely enjoy strong intellectual property protection. One well studied (and
reasonably well understood) situation where there are serious value capture problems is investment
in basic research and the production of scientific knowledge. Basic research usually ends up in sci-
entific publications, so it is hard e if not impossible e to secure strong intellectual property pro-
tection for scientific knowledge. As a result, it is very difficult to charge for discoveries, even if they
have the potential to generate high value for society, so very few firms invest in basic research. Spill-
overs (externalities) are simply too large; profiting from discovery is simply too difficult. There is no
easy for-profit business model for capturing value from scientific discoveries in a world where sci-
ence wants to be open and rapid dissemination of scientific knowledge through journals, confer-
ences and professional contacts is almost inevitable: not surprisingly, most basic research is not
funded by business firms, but by governments.

Investment in scientific research is an example of what economists call ‘public goods’; a circum-
stance in which the economic activity in question generates positive externalities or ‘spill-overs’. As
there is no good (private) business model that can support value capture, government funding and/
or philanthropy is required and provided. Viewed in this way, the concept of the ‘business model’
can be integrated into almost a century of economic thought about the design of institutions and
the role of enterprise and government in civil society. Market ‘failures’ occur in the context of
innovation when private business models for capturing value draw forth insufficient investment
in R&D.

Putting basic science to one side, the most common business model to capture value from in-
ventions is to embed them in a product, rather than simply trying to sell designs or intellectual
property. This approach allows those that invest in R&D to ameliorate (to some degree) their
lack of intellectual property protection. The latest cell phone, digital camera or automobile doesn’t
come with a price for the product and an unbundled price for knowhow and/or intellectual prop-
erty: invention/technology and product are typically bundled together, although (in theory) they
don’t need to be.

This discussion makes it apparent that market failures (with respect to R&D investment) are
partly a function of the ability (or lack thereof) of entrepreneurs to create viable business models
using the mechanisms available to them. As noted, one way to try and get around market failures in
the ‘market for inventions’ is to bundle invention(s) and complements into products. But too often,
firms (and in particular small start-ups) under-employ the available mechanisms, just offering cus-
tomers ‘items’ of technology such as devices or discrete technology components. Just by itself, this
may not represent a customer solution; a business model based on simply selling an invention e or
even an innovative component or ‘item’ e may not enable the innovator to capture a significant
share of the value that might be generated by their innovative technology, unless it has ironclad
patent protection and is critical to an important and already recognized application. The proper
‘marketing’ of new technology often requires much more.25 The bundled provision of complemen-
tary products and services is often necessary, not just to help capture value, but to help create it in
the first place.

The problem is quite general. When value delivery involves employing intangible (knowhow) as-
sets, pricing and value capture are difficult because of the nonexistence of perfect property rights,
which means that markets can’t work well, as Coase and many others have explained.26 As

Long Range Planning, vol 43 2010 185

illustrated above, many Internet services are simply provided for ‘free’ as a way to build brand and
to indirectly promote a related value added service, and we have seen how a mixture of revenue
approaches is usually required when trying to sell on the Internet.27 But bundling, while a common
and helpful approach, isn’t always necessary. When the innovator has a strong patent, it is some-
times possible to capture value either by naked licensing e or even outright sale e of intellectual
property. Different models of value capture are available where intellectual property rights exist and
can be enforced e so designing business models often requires the skill of the intellectual property
lawyer as well as that of the entrepreneur.

To summarize: the traditional revenue model used by innovators to capture value from technol-
ogy involves the consumer buying (and paying for) products that have intellectual property embed-
ded within them e the method is so common that it is rarely noticed or reflected on.28 This works
well, particularly if an attractive bundled solution can be offered, if there is strong intellectual prop-
erty, or if imitability is otherwise difficult. Many scientific discoveries and inventions are poorly
protected by intellectual property rights, and require business models that feature public funding,
or crafty ways to otherwise capture positive spill-overs.

business models innovation may not seem heroic [but] without it there

may be no reward for pioneering individuals, enterprises and nations.

Business models as innovation
Technological innovation is lionized in most advanced societies; that is a natural and desirable re-
flection of the values of a technologically progressive society. However, the creation of new orga-
nizational forms (like the Skunk Works and the multidivisional organizational structure),
organizational methods (like the moving assembly line), and in particular new business models
are of equal e if not greater e importance to society, and to the business enterprise. While such
innovation may seem less heroic to many citizens e even to many scientists and engineers e with-
out it technological innovation may be bereft of reward for pioneering individuals, as well as for
pioneering enterprises and nations.

The capacity of a firm (or nation) to capture value will be deeply compromised unless the capacity
exists to create new business models. As noted, even an inventor as celebrated as Thomas Edison had
a questionable track record in terms of business model innovation, abandoning the recording business
and also failing to get direct (rather than alternating) current adopted as the industry standard for elec-
tricity generation and transmission. History shows that, unless they can offer compelling value prop-
ositions to consumers/users and set up (profitable) business systems to satisfy them with the requisite
quality at acceptable price points, the innovator will fail, even if the innovation itself is remarkable, and
goes on to be widely adopted by society. Of course, this makes management, entrepreneurship and
business model design and implementation as important to economic growth as is technological in-
novation itself. Technological creativity that is not matched by business resourcefulness and creativity
(in designing business models) may not yield value to the inventor or even to their society.

As discussed and illustrated in many earlier examples, technological innovation often needs to be
matched with business model innovation if the innovator is to capture value. There are of course
exceptions e for example, small improvements in the manufacturing process (even if cumulatively
large) will usually not require business model innovation, and value can be captured by lowering
price and expanding the market and market share. But the more radical the innovation, and the
more challenging the revenue architecture, the greater the changes likely to be required to tradi-
tional business models. And, as indicated by some of the earlier examples, business model innova-
tion may help to establish a differentiable competitive advantage. Dell didn’t bring any
improvements to the technology of the Personal Computer e but it did combine both suppliers’
and its own organizational/distribution system innovations to deliver compelling value to end

186 Business Models, Business Strategy and Innovation

users: as have Southwest Airlines, Virgin, Virgin Blue, and JetBlue in the air passenger transport
sector.

Sometimes the creation of new business models leads to the creation of new industries. Consider the
payment card industry (the core of which is credit and debit cards). The card companies provide net-
work services, associate with banks who issue the cards, and associate with acquirers who sign up mer-
chants to accept credit cards. Early on in the life of the industry, merchants were unwilling to accept
a payment card that few consumers carried, just as card holders didn’t want cards that merchants did
not accept. As Evans and Schmalensee note, inventing a new business model for credit e the credit
card e ‘required the industry’s founders to invest enormous amounts of capital and ingenuity’.29

Companies should be seeking and considering improvements to business models e particularly
difficult to imitate improvements that add value for customers e at all times. Changing the firm’s
business model literally involves changing the paradigm by which it goes to market, and inertia is
likely to be considerable. Nevertheless, it is preferable for the firm to initiate such a change itself,
rather than have it dictated by external events, as several investment banks in the U.S. and elsewhere
have experienced recently.

The role of discovery, learning and adaptation
Designing a new business model requires creativity, insight, and a good deal of customer, compet-
itor and supplier information and intelligence. There may be a significant tacit component. An en-
trepreneur may be able to intuit a new model but not be able to rationalize and articulate it fully; so
experimentation and learning is likely to be required. As mentioned earlier, the evolving reality im-
pacting customers, society, and the cost structure of the business must be understood. It is often the
case that the right business model may not be apparent up front, and learning and adjustments will
be necessary: new business models represent provisional solutions to user/customer needs proposed
by represent entrepreneurs/managers. As Shirky recognizes, a business model is provisional in the
sense that it is likely over time to be replaced by an improved model that takes advantage of further
technological or organizational innovations. The right business model is rarely apparent early on in
emerging industries: entrepreneurs/managers who are well positioned, who have a good but not
perfect business model template but who can learn and adjust, are those more

likely to succeed

.30

The right business model is rarely apparent early on. entrepreneurs/

managers who are well positioned and can learn and adjust are more

likely to succeed

Technological change often provides the impetus for new and better ways to satisfy customer
needs. The horse, then the railroad, the auto and the airplane have all been technological solutions
to society’s basic transport needs that successively complemented and displaced each other, and
formed the basis of competing business models for carrying people from one place to another.
The Internet and the communication and computer revolution have empowered customers, and
both allowed and required more differentiation in product service offerings. Social networking is
also trumping the age-old ability of using advertising to get to an audience. As Peter Sealey has
noted with respect to new movies releases, ‘the star-power opening is fading in importance and the
marketing and releasing of movies is going into new territory where the masses are molding the opinion
of a movie’,31 and studio executives are having to recognize these new realities and adjust their busi-
ness models accordingly.

In short, one needs to distil fundamental truths about customer desires, customer assessments,
the nature and likely future behavior of costs, and the capabilities of competitors when designing
a commercially viable business model. Traditional market research will not often be enough to

Long Range Planning, vol 43 2010 187

identify as yet unarticulated needs and/or emerging trends. Changes with respect to the relative
merits of particular organizational and technological solutions to customer needs must also be
considered.

Consider again the question of how society will gather and distribute the news of the day. First it
was the town crier; later the newspaper; today the Internet has become increasingly important.
Communication costs have dropped dramatically; but now advertising revenues are shrinking
too. Generally, when the underlying technology changes, and an established logic for satisfying con-
sumer needs e e.g. newspapers for providing news e is overturned, the business model must
change too. But technological change is not always a trigger e or always necessary e to reshaping
the business model.

Not surprisingly, the invention of new business models can originate from many potential sour-
ces. What business models pioneers often possess e or develop e is an understanding of some
‘deep truth’ about the fundamental needs of consumers and how competitors are or are not satis-
fying those needs, and of the technological and organizational possibilities (and trajectories) for im-
provement e some of them, though, just stumble into such understandings. In almost every case,
however, a new business model is successfully pioneered only after considerable trial and error.

Those entrepreneurs who understand ‘deep truths’ and can figure out what customers want and
design a better way to satisfy them (and build sustainable organizations to address these customer
needs) are business pioneers. They may or may not use new technology, but they must understand
customer needs, technological possibilities, and the logic of organization. Put differently, a business
model articulates the underlying business or ‘industrial logic’ of a firm’s go-to-market strategy.
Once articulated, it is likely that the logic will have to be tested and retested, adjusted and tuned
as the evidence with respect to provisional assumptions becomes clarified.

Netflix (discussed above), the largest online DVD rental service in the U.S., offers a flat-fee DVD
movie rental service that, by 2007, was serving over 6 million subscribers from its collection of
75,000 titles.32 Subscribers can use the website’s browse function to search for movies by genre,
and use an extensive movie recommendation system based on other users’ ratings to add to their
ordered list for delivery via mail. At its initial launch, the Netflix business model was based on
a pay-per-rental service, but this initial pricing model did not succeed, and the company almost
failed. It was clear to management Netflix had to rejig its business model and, between September
and October 1999, it reinvented itself with a subscription model (the ‘Marque Program’). It ended
its pay-per-rental model entirely, and evolved the monthly fee program to allow subscribers to rent
any number of DVDs per month (although only a limited number at any one time). The model was
supported by a system of regional distribution centers which ensured next day delivery to over 90%
of subscribers. Clearly, it took a while to be able to ascertain the right price points and the manner
of pricing that was most acceptable to the customer base for its new service; but as Netflix manage-
ment figured out viewer convenience, wants and willingness to pay, it adjusted its business model
accordingly. This ability to perceive and adapt saved Netflix and laid the foundation for its growth
and development: by 2006 it had reached almost $1 billion in revenues.

Selecting the right ‘architecture’ and pricing model for a business requires not just understanding
the choices available, but also assembling the evidence needed to validate conjectures and hunches
about costs, customers, competitors, complementors, distributors and suppliers takes detailed fact-
specific inquiry, and a keen understanding of customer needs and customer willingness to pay, as
well as of competitor positioning and likely competitive responses. Entrepreneurs and executives
must make many informed guesses about the future behavior of customer and competitor, as
well as of costs. As the evidence with respect to initial conjectures becomes available, they need
to adjust accordingly. Being fast in learning and making the requisite adjustments to the model
is important.

A helpful analytic approach for management is likely to involve systematic deconstruction/un-
packing of existing business models, and an evaluation of each element with an idea toward refine-
ment or replacement. The elements of a business model must be designed with reference to each
other, and to the business/customer environment and the trajectory of technological development

188 Business Models, Business Strategy and Innovation

in the industry. While the questions are not as crisp as one would like, and the answers are likely to
be ambiguous, endeavoring to answer them will impose some discipline and at least help one sort
business propositions that are likely to be viable from those that are not. For instance, business
propositions that are no more than good ideas fall short; likewise propositions that involve captur-
ing 1% of huge markets show a lack of understanding of differences amongst (potential) customers,
market segments and competition. And wonderfully novel (gimmicky) product concept that meets
the needs of but a handful of potential customers is unlikely to yield much value. Periodic review
can increase the chances of avoiding blind spots: long-lived structural elements e choices made per-
haps decades ago in different environments e need to be scrutinized especially thoroughly.

A provisional business model must be evaluated against the current

state of the business ecosystem, and against how it might evolve

A provisional business model must be evaluated against the current state of the business ecosys-
tem, and also against how it might evolve. Questions to consider (which are summarized in
Figure 3) include:

� How does the product or service bring utility to the consumer? How is it likely to be used? In-
asmuch as innovation requires the provision of complements, are the necessary complements al-
ready available to the consumer with the convenience and price that is desirable (or possible)?;

� What is the ‘deep truth’ about what customers really value and how will the firm’s service/prod-
uct offering satisfy those needs? What might the customer ‘pay’ for receiving this value?;

� How large is the market? Is the product/service honed to support a mass market?;
� Are there alternative offerings already in the market? How is the offering superior to them?;
� Where is the industry in its evolution? Has a ‘dominant design’ emerged? Strategic requirements

are likely to be different in the pre- and post-paradigmatic periods;33

� What are the (contractual) structures needed to combine the activities that must be performed to
deliver value to the consumer? Both lateral and vertical integration and outsourcing issues need
to be considered. (Contract theory/transaction cost economics is a useful lens through which to
view many of these issues. So is capability theory);

� What will it cost to provide the product/service? How will those costs behave as volume and
other factors change?; and

� What is the nature of the appropriability regime? How can imitators be held at bay, and how
should value be delivered, priced, and appropriated?

As the author has noted in previous work, beyond specifying a realistic revenue architecture, de-
signing a business model also involves determining the set of lateral (complementary) and vertical
activities that must be performed and assessing whether and how they can be performed sufficiently
cheaply to enable a profit to be earned, and who is to perform them. It involves figuring out the
market entry strategy e while entry timing is a strategic, rather than a business model issue, it
may depend in part on the business model employed, particularly the complements already in
place.34

When establishing a new business there is likely to be uncertainty with respect to all of the above.
Disappointments are certain to arise as a new business is built, but success rates can be improved if
the architects of the business model learn quickly, and are able to adjust within a range that still
yields a satisfactory profit.

Of course, once a business model is successfully established, changing technology and enhanced
competition will require more than defenses against imitation. It is also likely that even successful
business models will at some point need to be revamped, and possibly even abandoned. For exam-
ple, as the value proposition associated with the traditional personal computer software licensing

Long Range Planning, vol 43 2010 189

Figure 3. Questions to ask about a (provisional) business model

model (whereby periodic updates would require the purchase of new software licenses and addi-
tional maintenance costs) has weakened for some customers, Microsoft has changed elements of
its business model to allow renting so as to compete with cheap or free Web alternatives. According
to one source, Microsoft is ‘overhauling not only what it makes but how to deliver and charge for it’.35

Microsoft has apparently begun to offer its Exchange email server program for a monthly fee, as
well as a ‘barebones’ version of Office for free, supported in part by online advertising (in fact,
it appears now to be offering some products under the ‘freemium’ philosophy described earlier).
The evidence is not yet in as to whether it will work well for both Microsoft and its customers.

designing good business models is an ‘art’ .. the chances are greater if

entrepreneurs and managers have a deep understanding of user needs

and are good listeners and fast learners.

Clearly, designing good business models is in part an ‘art’. The chances of good design are greater
if entrepreneurs and managers have a deep understanding of user needs, consider multiple alterna-
tives, analyze the value chain thoroughly so as to understand just how to deliver what the customer
wants in a cost-effective and timely fashion, adopt a neutrality or relative efficiency perspective to
outsourcing decisions, and are good listeners and fast learners. Useful tools include the various
types of market research that lead to a deep understanding of the user, along with elements of
the Profiting from Innovation framework such as the innovation cycle, appropriability regimes,
complementary assets and intellectual property systems.

The selection/design of business models is a key microfoundation of dynamic capabilities e the
sensing, seizing, and reconfiguring skills that the business enterprise needs if it is to stay in synch
with changing markets,36 and which enable it not just to stay alive, but to adapt to and itself shape
the (changing) business environment. Dynamic capabilities help govern evolutionary fitness, and

190 Business Models, Business Strategy and Innovation

help shape the business environment itself. Get the business model wrong, and there is almost no
chance of business success e get it right, and customize it for a market segment and build in non-
imitable dimensions, and it will contribute to the firm’s competitive advantage.

Magretta claims that business models are ‘variations on the generic value chain underlying all busi-
nesses’. This view would seem to overlook that a business model is only partly about how to orga-
nize the value chain e it is also about figuring out the value proposition to the customer as well as
the value capture mechanism. A sustainable business model is as much (as the current author has
noted) about where to position within the value chain i.e. what are the key bottleneck assets to own/
control in order to capture value. Clearly, the industry must perform various activities in the value
chain e but which one(s) the firm chooses to undertake is very much a business model choice.

Recognized (but not fully developed here) is the notion that a business model cannot be assessed
in the abstract; its suitability can only be determined against a particular business environment or
context. Neither business strategies, business structures nor business models can be properly cali-
brated absent assessment of the business environment; and of course the business environment it-
self is, in part, a choice variable; i.e. firms can both select a business environment, and be selected by
it: and they can also shape their environment.

.the business environment itself is a choice variable: firms can select

a business environment or be selected by it: they can also shape it

Zott and Amit bravely endeavor to hypothesize as to the appropriate mapping of business models
to two product market choices: cost leadership and differentiation.37 However, our state of under-
standing as to the precise relationship between business model choice and enterprise performance is
both highly context dependent and rather primitive. In certain contexts (e.g. market entry strategies
for innovators) testable propositions have been advanced (including by the current author), but
strategic studies will have to advance further as a field before mapping can be anything other
than suggestive.

Of course, it may very well take time to get a business model right. Pioneers, in particular, are
often forced to make only educated guesses as to what customers want, what they will pay for,
and the cost structures associated with various ways to organize. As the author’s ‘Profiting from
Innovation’ paper discusses, especially in the pre-paradigmatic industry evolution phase, it is nec-
essary to stay flexible, experiment with the product and the business model and learn, both from
one’s own and one’s competitors’ activities, and to keep sufficient financial resources on hand to
remain an industry participant e and hopefully the market leader e by the time the ‘dominant de-
sign’ emerges in the market. Indeed, one hopes to be the promoter/owner of this dominant
design e and to have the capacity to capitalize on the situation.

Conclusion
All businesses, either explicitly or implicitly employ a particular business model. A business model de-
scribes the design or architecture of the value creation, delivery and capture mechanisms employed. The
essence of a business model is that it crystallizes customer needs and ability to pay, defines the manner by
which the business enterprise responds to and delivers value to customers, entices customers to pay for
value, and converts those payments to profit through the proper design and operation of the various el-
ements of the value chain. Put differently, a business model reflects management’s hypothesis about what
customers want, how they want it and what they will pay, and how an enterprise can organize to best meet
customer needs, and get paid well for doing so. The goal of this article has been to advance understanding
of the considerable significance of business models and to explore their connections to business strategy,
innovation management and economic theory.

Long Range Planning, vol 43 2010 191

One key conclusion of the analysis is that, to be a source of competitive advantage, a business
model must be something more than just a good logical way of doing business. A model must
be honed to meet particular customer needs. It must also be non-imitable in certain respects, either
by virtue of being hard to replicate, or by being unpalatable for competitors to replicate because it
would disturb relationships with existing customers, suppliers, or important alliance partners. A
business model may be difficult for competitors to replicate for other reasons too. There may be
complicated process steps or strong intellectual property protection, or organizational structures
and arrangements may exist that will stand in the way of implementing a new business model.
Good business model design and implementation involves assessing such internal factors as well
as external factors concerned with customers, suppliers, and the broader business environment.

to be a source of competitive advantage, a business model must be

more than just a good logical way of doing business .. It must be

honed to meet particular customer needs .

The paucity of literature (both theoretical and practical) on the topic is remarkable, given the
importance of business design, particularly in the context of innovation. The economics literature
has failed to even flag the importance of the phenomenon, in part because of an implicit assump-
tion that markets are perfect or very nearly so. The strategy and organizations literature has done
little better. Like other interdisciplinary topics, business models are frequently mentioned but rarely
analyzed: therefore, they are often poorly understood. Not surprisingly, it is common to see great
technological achievements fail commercially because little, if any, attention has been given to de-
signing a business model to take them to market properly.

This can and should be remedied. Increased understanding of the essence of business models and
their place in the corpus of the social and organizational sciences should help our understanding of
a variety of subjects including market behavior, competition, innovation, strategy and competitive
advantage. Our understanding of the nature of the firm itself, together with the role of entrepre-
neurs and managers in the economy and in society, should also benefit from a better appreciation
of business models and their role in entrepreneurship, innovation and business performance.

great technological achievements commonly fail commercially

because little attention has been given to designing a business model

to take them to market properly. This can and should be remedied.

Acknowledgement
I would like to thank Charles Baden-Fuller and Ian MacMillan for their invitation to contribute to this
Special Issue, and Michael Akemann, Sebastien Belanger, John Blair, Hank Chesbrough, Michael Katz,
Doug Kidder, David Mitchell, Charles O’Reilly, Richard Rumelt, Alexander Stern, Leigh Teece and
Steve Lewis as well as the principals of Living PlanIT SA for helpful insights into the issues discussed
here. The skilful assistance of Patricia Lonergan in preparing the manuscript is gratefully acknowledged.

References

1. There are other (related) definitions of a business model. Amit and Zott see R. Amit and C. Zott, Value

creation in e-business, Strategic Management Journal 22, 493e520 (2001); and C. Zott and R. Amit, The fit

192 Business Models, Business Strategy and Innovation

between product market strategy and business model: implications for firm performance, Strategic Man-
agement Journal 29, 1e26 (2008) define a business model as ‘the structure, content, and governance of trans-
action’ between the focal firm and its exchange partners (e.g. customers, vendors, complementors). For yet
another alternate definition see Chesbrough and Rosenbloom (see following).

2. H. Chesbrough and R. S. Rosenbloom, The role of the business model in capturing value from innova-
tion: evidence from xerox corporation’s technology, Industrial and Corporate Change 11(3), 529e555
(2002).

3. The end of the free lunch e again, The Economist 390(8623) (March 21st 2009).
4. See K. Arrow, The Limits of Organization, Norton, New York, (1974). The ArroweDebreu model of com-

petitive equilibrium has everything priced; but, as Arrow himself notes elsewhere, ‘in a strictly technical
and objective sense, the price system does not work. You simply cannot price certain things (p. 22) and ‘ trust
and similar values, loyalty and truth telling e are not commodities for which trade in the open market is tech-
nically possible or even meaningful. (p. 23). ‘ A firm. provides another major area within which price
relations are held in partial abeyance. (p. 25).

5. The structure-conduct-performance paradigm in the field of industrial organization is possibly an excep-
tion. It stressed that concentrated markets were more profitable. If translated into management/strategy
nostrums, as Michael Porter, Competitive Strategy, Free Press, (1982) did, it suggest the benefits of either
scale or differentiation as profit drivers. While scale and differentiation may still assist as profit drivers, the
situation in the modern economy is that in many circumstances, these nostrums can be quite misleading.

6. O. E. Williamson, Organizational innovation: the transaction-cost approach (1983), in J. Ronen (ed.),
Lexington Books, Lexington, MA (1983).

7. R Miles, G. Miles, C. Snow, K. Blomquist and H. Rocha, Business Models, Organizational Forms, and
Managerial Values, Working paper, UC Berkeley, Haas School of Business (2009). The authors note
how new business models, new organizational forms, new management approaches, and entrepreneurship
are the foci of different groups of scholars who rarely meet.

8. G. Porter, The Rise of Big Business, 1860e1910, Harland Davidson, Arlington Heights, Illinois, (1973) p. 49.
9. C. W. Ebeling, Evolution of a box: the invention of the intermodal shipping container revolutionized the

international transportation of goods, Invention and Technology 8e9 (2009).
10. Apple’s iTunes music store is an example of a business model innovation, and was the first legal pay-

as-you-go method for downloading music. Time Magazine hailed it as ‘ the coolest invention for 2003’.
11. See C. Shapiro and H. Varian, Information Rules: A Strategic Guide to the Network Economy, Harvard Busi-

ness School Press, Boston, MA, (1999) The rules for strategic engagement that they promulgate are core
elements of strategy in the information services sector, and here e as elsewhere e the design of business
models to support sustainable competitive advantage must be informed by strategy analysis.

12. A. Shuen, Web 2.0: A Strategy Guide, O’Reilly, Sebastopol, (2008) p. 2.
13. See also J. B. Harreld, C. A. O’Reilly and M. L. Tushman, Dynamic capabilities at IBM: driving strategy

into action, California Management Review 49(4) (2007).
14. M. Dell, The Early Entrepreneurial Years in Starting a Business, Harvard Business School Press, (2008) In-

deed, a critical element of Dell’s success is not just the way it has organized the value chain, but also the
products that it decides to sell through its distribution system. The initial products were personal
computers, but now include printers, digital projectors, and computer-related electronics.

15. Quoted in J. Magretta, Why business models matter, Harvard Business Review 6 (2002).
16. For an insightful treatment of the Google story, see D. A. Vise, The Google Story, Bantam Dell, New York

(2008).
17. See J. M. de Figueiredo and D. J. Teece, Mitigating Procurement hazards in the context of innovation,

Industrial and Corporate Change 5(2), (1996) for an analysis of some ways to mitigate the hazards of
competing with one’s suppliers.

18. S. Lippman and R. Rumelt, Uncertain imitability: an analysis of interfirm differences in efficiency under
competition, Bell Journal of Economics 13, 413e438 (1982).

19. D. J. Teece, Profiting from technological innovation: implications for integration, collaboration, licensing
and public policy, Research Policy 15(6), 285e305 (1986); D. J. Teece, Reflections on profiting from tech-
nological innovation, Research Policy 35(8), 1131e1146 (2006).

20. See D. J. Teece (1986) ibid.; and G. Pisano and D. J. Teece, How to capture value from innovation: shap-
ing intellectual property and industry architecture, California Management Review 50(1), 278e296 (2007).

21. D. J. Teece, The multinational enterprise: market failure and market power considerations, Sloan Manage-
ment Review 22(3), 3e17 (1981).

Long Range Planning, vol 43 2010 193

22. S. Winter, The logic of appropriability: from Schumpeter to Arrow to Teece, Research Policy 35,
1100e1106 (2006).

23. D. J. Teece, Explicating dynamic capabilities: the nature and microfoundations of (sustainable) enterprise
performance, Strategic Management Journal 28(13), 1319e1350 (2007).

24. An application of the framework in the biotech industry context is discussed in G. Pisano, Science Business:
The Promise, the Reality, and the Future of Biotech, Harvard Business School Press, (2006) which includes
a carefully analysis of the sources of failure in the market for know how.

25. For further development of this idea, see W. Davidow, High Technology Marketing, Free Press (1986).
26. R. Coase, The problem of social cost, Journal of Law and Economics 3, 1e44 (1960).
27. The proliferation of illegal digital downloads of recorded music has led recording companies to try to in-

sist on (and sometimes achieve) so called ‘360 contracts’, so they can participate in all sources of revenue
from their artistes’ activities e branded clothing, performances, and other public appearances e as well as
just recorded music.

28. D. J. Teece (1986) op. cit. at Ref 19; D. Somaya and D. J. Teece, Patents, licensing and entrepreneurship:
effectuating innovation in multi-invention contexts in Sheshinki, Baumol and Strom (eds.), Entrepreneur-
ship, Innovation, and the Growth of Free-Market Economies, Princeton University Press (2007).

29. D. Evans and R. Schmalensee, Paying with Plastic, MIT Press, Cambridge, (1999) p. 3.
30. C. Shirky, Here Comes Everybody: The Power of Organizing Without Organizations, Penguin, New York

(2008).
31. See Claudia Eller ‘Little Love this Summer for A-List Movie Actors’, Los Angeles Times, 29 June, 2009.
32. http://ir.netflix.com (visited April 2007).
33. See D. J. Teece (1986) op. cit. at Ref 19.
34. W. Mitchell and Dual Clocks, Entry order influences on industry incumbents and newcomer market share

and survival when specialized assets retain their value, Strategic Management Journal 12(2), 85e100
(1991).

35. Peter Burrows, Microsoft defends its empire, Business Week p. 28 (6 July 2009).
36. D. J. Teece, G. Pisano and A. Shuen, Dynamic capabilities and strategic management, Strategic Manage-

ment Journal 18(7), 509e533 (1997); D. J. Teece (2007), op. cit. at Ref 23; D. J. Teece, Dynamic Capabil-
ities and Strategic Management: Organizing for Innovation and Growth, Oxford University Press (2009).

37. C. Zott and R. Amit (2008), op. cit. at Ref 1.

Biography
David J. Teece has a Ph.D. in economics from the University of Pennsylvania. His research interests span

industrial organization, business strategy, organizational economics, and public policy. He is the author of over

200 published articles and books. His most recent book is Dynamic Capabilities and Strategic Management:

Organizing for Innovation and Growth (Oxford University Press, 2009). He has four honorary doctorates and

was the co-founder and Vice Chairman of LECG Corporation. Institute for Business Haas School of

Business University of California, Berkeley Berkeley, California 94720. Tel: 510-642-1075; Fax: 510-642-2826;

E-mail: davidjteee@teece.net

194 Business Models, Business Strategy and Innovation

http://ir.netflix.com

http://davidjteee@teece.net

Business Models, Business Strategy and Innovation
Introduction
Business models – the theoretical foundation
Examples of business models
Traditional industries
The information/internet industries
Business models, strategy and sustainable competitive advantage
Barriers to imitating business models
Business model learning
Business models to capture value from technological innovation
The profiting from innovation framework
‘Public’ goods and the bundling and unbundling of inventions and products
Business models as innovation
The role of discovery, learning and adaptation
Conclusion
Acknowledgements
References

1/23/2020

Business Strategy and Forecasting as Competitive Advantages Scoring Guide

https://courserooma.capella.edu/bbcswebdav/institution/BMGT/BMGT8130/200100/Scoring_Guides/u02a1_scoring_guide.html 1/1

Business Strategy and Forecasting as Competitive Advantages Scoring Guide

Due Date: End of Unit 2
Percentage of Course Grade: 30%.

CRITERIA NON-PERFORMANCE BASIC PROFICIENT DISTINGUISHED

Describe a business
model, using Teece
(2010) to analyze the
strategy of the
organization.
20%

Does not
explain the
background of
the business as
related to its
strategic
business model,
and/or does not
cite Teece
(2010).

Provides the
background of a
business and its
model, but does not
align with Teece
(2010), or fails to
provide strategic
insights.

Describes a business
model, using Teece
(2010) to analyze the
strategy of the
organization.

Analyzes a business
model, using Teece (2010)
to evaluate the strategy of
the organization.

Describe the
business strategy
using market
segmentation, value
proposition,
apparatus, and
prevention of
imitability, and
describe the
competitive
advantage of the
organization.
Substantively use
Teece (2010) as
support.
20%

segmentation,
value
proposition,
apparatus, or
prevention of
imitability, and
does not
describe the
competitive
advantage of
the
organization.

Discusses some of
the business strategy
using market
segmentation, value
proposition,
apparatus, or
prevention of
imitability. Does not
cite Teece (2010) for
substantive
purposes.

Describes the business
strategy using market
segmentation, value
proposition, apparatus,
and prevention of
imitability, and describes
the competitive
advantage of the
organization.
Substantively uses Teece
(2010).

Evaluates the business
strategy using market
segmentation, value
proposition, apparatus, and
prevention of imitability,
and describes the
competitive advantage of
the organization. Uses
Teece (2010) to support the
evaluation.

Analyze and
describe Part C1 of
the assignment,
using one of the Red
Queen articles as
support.
20%

Does not
incorporate Part
C1 of the
assignment.

Describes Part C1 of
the assignment but
fails to explain or
incorporate the Red
Queen literature or
leaves out some
parts.

Analyzes and describes
Part C1 of the
assignment, using one of
the Red Queen articles
as support.

Thoroughly evaluates,
analyzes, and describes
Part C1 of the assignment,
using both Red Queen
articles as support.

Describe the Red
Queen articles.
Show some
differences between
the articles.
20%

Does not
analyze,
evaluate, or
synthesize the
Red Queen
articles.

Recites the
information in the
Red Queen articles.

Describes the Red
Queen articles. Shows
some differences
between the articles.

Synthesizes the Red
Queen articles thoroughly.
Explains how the concept
has grown in the past
decade.

Communicate in a
manner expected of
doctoral-level
composition and
exhibit critical
thinking skills.
20%

Does not
communicate in
a manner
expected of
doctoral-level
composition
and does not
exhibit critical
thinking skills.

Inconsistently
communicates in a
manner expected of
doctoral composition
and inconsistently
exhibits critical
thinking skills.

Communicates in a
manner expected of
doctoral-level
composition and exhibits
critical thinking skills.

Communicates in a manner
expected of doctoral-level
composition, follows APA
conventions of writing with
few to no errors, and
exhibits exceptional critical
thinking skills.

THE RED QUEEN EFFECT: COMPETITIVE ACTIONS AND
FIRM PERFORMANCE

PAMELA J. DERFUS

PATRICK G. MAGGITTI
Temple University

CURTIS M. GRIMM
KEN G. SMITH

University of Maryland

We investigate the Red Queen effect as a contest of competitive moves or actions among
rivalrous firms. The results from a multi-industry study of over 4,700 actions confirms
the existence of Red Queen competition, whereby a firm’s actions increase perfor-
mance but also increase the number and speed of rivals’ actions, which, in turn,
negatively affect the initial firm’s performance. We further show that this Red Queen
effect depends on industry context and a focal firm’s market position.

The quest to explain performance differences
among competing firms is a fundamental issue in
strategic management. A number of answers to this
complex question have been offered. According to
the industry structure viewpoint, positioning firms
in industries where they can take advantage of fa-
vorable competitive forces, such as barriers to entry
or mobility (Caves & Porter, 1977), enhances per-
formance. The resource-based view also empha-
sizes limiting the behavior of rivals by suggesting
that firms acquire or develop unique, valuable, and
rare resources that are difficult for rivals to repli-
cate (Barney, 1986). Evolutionary theory posits per-
formance differences among firms are a function of
a competitive race to discover profit opportunities.
According to this view, high performance is
achieved by speed and innovation that keep firms
ahead of rivals (Nelson & Winter, 1982). Our focus
in this paper is the latter perspective, perhaps the
least understood of the three; more specifically, we
explore “Red Queen competition” in the context of
actions among rivals.

Evolutionary and ecology theories focusing on
Red Queen competition portray how entities dy-
namically interact and coevolve with one another.
Introduced by the biologist van Valen (1973), the

Red Queen effect is based on the conversation be-
tween the Red Queen and Alice in Lewis Carroll’s
Through the Looking Glass. In that story, Alice
realizes that although she is running as fast as she
can, she is not getting anywhere, relative to her
surroundings. The Red Queen responds: “Here, you
see, it takes all the running you can do, to keep in
the same place. If you want to get somewhere else,
you must run at least twice as fast as that!” (Carroll,
1960: 345).1 Van Valen used this analogy to de-
scribe the continuous and escalating activity and
development of participants trying to maintain rel-
ative fitness in a dynamic system. Since then, the-
orists have used the notion of the Red Queen to
explain behavior in a variety of settings ranging
from biology to military arms races (Baumol, 2004;
Dawkins & Krebs, 1979).

Applied to a business context, the Red Queen can
be seen as a contest in which each firm’s perfor-
mance depends on the firm’s matching or exceed-
ing the actions of rivals. In these contests, perfor-
mance increases gained by one firm as a result of
innovative actions tend to lead to a performance
decrease in other firms. The only way rival firms in
such competitive races can maintain their perfor-
mance relative to others is by taking actions of their
own. Each firm is forced by the others in an indus-
try to participate in continuous and escalating ac-
tions and development that are such that all the
firms end up racing as fast as they can just to stand
still relative to competitors. This self-escalating,

This research was partially supported by the Dingman
Center for Entrepreneurship at the Robert H. Smith
School of Business, University of Maryland, College
Park. The authors would also like to thank our AMJ
associate editor, R. Duane Ireland, and the three anony-
mous reviewers for their insightful comments and sug-
gestions during the revision process.

1 Through the Looking Glass was originally published
in 1871.

� Academy of Management Journal
2008, Vol. 51, No. 1, 61–80.

61

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express
written permission. Users may print, download or email articles for individual use only.

coevolving system of Red Queen competition has
been empirically shown to affect founding rates
(Barnett & Sorenson, 2002), failure rates (Barnett &
Hansen, 1996), and competitiveness (Barnett &
McKendrick, 2004). Indeed, Baumol (2004) sug-
gested that the Red Queen effect is the most pow-
erful mechanism driving economic development in
capitalistic society.

Following Barnett and McKendrick (2004), we
build a model that captures the Red Queen process
of competitive evolution as both a positive and a
negative force on focal firm performance in which
the gains by one firm must come at the expense of
another. As Barnett and McKendrick noted, “A de-
fining characteristic of competition is that one or-
ganization’s solution becomes its rivals’ problem.
The resulting increased constraints, again in turn,
are likely to trigger responses among rivals, again
intensifying competitive constraints on the first or-
ganization, and so on” (2004: 540). The first pur-
pose of this study was to explicitly model the Red
Queen “running as fast as you can” process by
examining the relationships among focal firm ac-
tions, rival actions, speed of rival actions, and focal
firm performance. In doing this, we illustrate theo-
retically and empirically that focal firm actions ver-
sus rival actions and speed of rival actions have
two opposing effects on focal firm performance. We
refer to this formulation as our baseline model. A
second goal of our research was to expand the
baseline model by developing and testing theory
that begins to identify the conditions that moderate
the Red Queen effect.

Our examination of firm action and firm perfor-
mance and the coevolutionary dependency of firm
action and rival action on firm performance is fin-
er-grained and more dynamic than prior Red Queen
research. Specifically, we contend that firms are
prompted to search, undertake new actions, and
learn in an effort to improve performance. This use
of Red Queen theory is consistent with Schumpet-
er’s 1942 argument that a dynamic process of “cre-
ative destruction” occurs when firms launch inno-
vative actions to gain advantage in the marketplace,
which is then eroded by their rivals’ competitive
moves (Schumpeter, 1976). Thus, Schumpeter’s
perspective figures prominently in the develop-
ment of our Red Queen model. Additionally, our
study applies Red Queen theory in the context of
competitive dynamics, which focuses on the ac-
tions and reactions of firms (Smith, Grimm, & Gan-
non, 1992). In line with prior competitive dynam-
ics research, we define firm actions as “externally
directed, specific and observable competitive
moves initiated by a firm to enhance its relative
competitive position” (Smith, Ferrier, & Ndofor,

2001: 321). Rival actions are defined as the exter-
nally directed competitive moves of all rivals in the
industry in which the focal firm’s participation is
being studied (Young, Smith, & Grimm, 1996).

Although our conceptualization of Red Queen
competition is consistent with Schumpeterian and
competitive dynamics perspectives, neither
Schumpeter nor competitive dynamics research
fully explains the motivating factors behind the
competitive process. The Red Queen theory out-
lined in this article provides a more complete pic-
ture of this competitive process by explaining how
firms are motivated to search, act, and learn in a
desire to improve performance. As a consequence,
our Red Queen theory begins to clarify vital aspects
of the competitive process—the motivation for ac-
tion and reaction—not fully explained in prior
literature.

Like Barnett and Hansen (1996), we assume that
a firm facing competition is likely to act. Further,
we regard the competitive interactions of firms as
constituting a mechanism for a simple search, ac-
tion, and learning process firms undertake to im-
prove performance (March & Simon, 1958). In our
model, action allows firms to “learn by doing” (Ar-
gote, 1999; Eisenhardt & Tabrizi, 1995; Pisano,
1994). Actions of focal firms that increase their
performance may result in a decline in rival perfor-
mance, thus prompting those rivals to engage in
similar search, action, and learning processes. Our
goal was to model this reciprocal system of focal
firm actions and rival actions, including the speed
at which those rival actions occur, and to show the
system’s effect on focal firm performance. We focus
on short-run performance, or the effects of focal
firm action and rival action on focal firm perfor-
mance in a given year. Though we recognize that
Red Queen theory also posits long-term conse-
quences of action exchanges, such as greater fitness
for all competing firms, these long-term effects are
beyond the scope of the present work.

As mentioned above, in addition to examining
how firm actions, rival actions, and rival action
speed impact short-term performance, a second
goal of our research was to develop and test theory
that begins to identify the conditions that amelio-
rate or exacerbate the Red Queen effect. Such work
is important as it may offer insights into how firms
can successfully adapt or evolve under Red Queen
competitive pressures. In this research, we focus on
a set of factors that may facilitate or impede how
managers learn from action. We argue that a better
understanding of these factors can help explain the
motivation for search and action and thereby, Red
Queen competition. We suggest that industry con-
centration, industry growth rate, and a firm’s mar-

62 FebruaryAcademy of Management Journal

ket position affect the search and action process
and hence moderate the Red Queen effect.

RED QUEEN COMPETITION: FOCAL FIRM
ACTIONS, RIVAL ACTIONS, RIVAL ACTION
SPEED, AND FOCAL FIRM PERFORMANCE

In this section, we develop baseline theory that
explains the Red Queen effect in terms of focal firm
actions, rival actions, and rival action speed, and
their combined impact on focal firm performance.
We argue that search and firm action are motivated
by a desire to learn new ways of improving perfor-
mance. However, we assume that the relationship
between firm action and performance is uncertain,
dynamic, and subject to constant change by the
very Red Queen competition that motivates search
and action. In our model, firms search to discover
opportunities to act. They experiment in taking
new actions and learn from the results of action
about the relationship between action and perfor-
mance. We do not assume that all action is effective
or costless, but we do assume that the average per-
formance benefits of action outweigh the costs.
Otherwise, firms would not have a motivation
to act.

Recent theory on the Red Queen effect has spec-
ulated about the motivation for firms’ and rivals’
actions. As Barnett and McKendrick (2004) sug-
gested, firm aspirations expand, and goals may
change quickly as a result of comparisons with
others and Red Queen evolution. For instance, a
large retailer may increase the frequency of market-
ing campaigns on the basis of observing a positive
relationship between prior campaigns and perfor-
mance. Assume this action results in increased rev-
enue and profits for the focal firm at the expense of
a rival’s profits. Rivals may then search for and
learn of some way to increase their own perfor-
mance. In this example, rivals may take new price-
cutting actions. These rival actions may adversely
affect the performance of the focal firm, thus moti-
vating additional search, action, and learning for
this retailer. The learning from action in pursuit of
profits that drives this action–rival action process
captures the incremental and coevolutionary na-
ture of Red Queen competition.

The motivation to search, act, and learn elicited
by the Red Queen effect extends Schumpeter’s the-
ory of creative destruction regarding the relation-
ship between action and performance in a compet-
itive context. Schumpeter (1934) highlighted the
interdependent nature of a competitive market-
place, arguing that it was the result of, and the
reason for, continuous innovation and firm action.
If firms stand still, competitors who introduce new

combinations that appeal to the market erode the
inactive firms’ positions. To avoid this erosion,
firms must continually strive to introduce new
products, methods, and initiatives. Success, or
profitable performance, he argued, is more the re-
sult of “the new commodity, the new technology,
the new source of supply, the new type of organi-
zation” than it is the result of control of margins,
output, or prices (Schumpeter, 1976: 84 – 85). In
Schumpeter’s world, an efficient but lethargic firm
would not survive for long.

Schumpeter also pointed out, however, that in-
novation and action draw rival action, which he
referred to as “creative destruction.” The acting
firm “leads in the sense that he draws other pro-
ducers in his branch after him” (Schumpeter, 1976:
89). Thus, innovations and the results of new ac-
tion are visible to competitors, often spurring rival
actions and an ongoing cycle of creation and de-
struction. Indeed, if rivals are to survive in the
marketplace, they cannot afford to ignore the com-
petitive actions of other firms; they must also act
creatively. The innovative competitive interaction
of firms in pursuit of profits is so fundamental that
Schumpeter (1976) argued it was the key source of
market expansion and economic growth.

Researchers have empirically found that firms
that are more active (i.e., are running faster) than
their rivals improve their competitive positions
(Ferrier, Smith, & Grimm, 1999) and increase their
performance (Young, 1993; Young et al., 1996),
while firms that are more sluggish than their rivals
experience negative performance consequences
(Miller & Chen, 1994). The basic argument of this
research has been that more active firms achieve
greater performance because they have greater as-
piration levels, are more capable at implementing
actions, and are perceived by rivals as more aggres-
sive competitors than are less active firms (Smith et
al., 2001).

Yet neither Schumpeter (1976) nor competitive
dynamics researchers have recognized the role that
learning from action outcomes may play in fueling
competition and evolution. Barnett and McKen-
drick (2004) described how learning drives Red
Queen competition. When performance falls below
aspirations, managers will search, act, and learn
until performance reaches expectations. At the
heart of this process is a manager seeking to under-
stand a dynamic world of action cause and effect.
Weick noted that managers cannot “ignore the ac-
tion because they are responsible for it” (1995:
134 –135). Thus, Red Queen theory helps to explain
how firms incrementally evolve by taking action
and learning from the results of action in a desire to

2008 63Derfus, Maggitti, Grimm, and Smith

improve performance. Given these arguments along
with prior research, we predict:

Hypothesis 1a. With the number of rival ac-
tions held constant, as the number of focal firm
actions increases, focal firm performance
increases.

However, Red Queen competition can also have
negative consequences for an active firm (Barnett &
Hansen, 1996; Barnett & McKendrick, 2004). Bar-
nett and McKendrick contended that one organiza-
tion’s solution to search and action can become
another firm’s problem. In this sense, Red Queen
competition narrows the options for firms, escalat-
ing rivalry and races, sometimes with limited short-
term benefits for all. Returning to the above exam-
ple, consider a case in which a rival’s response was
not to cut prices but to simply imitate an initially
acting firm’s behavior by increasing the frequency
of its own marketing campaigns, thereby regaining
the revenues shifted by the focal firm’s initial
action.

Schumpeter argued that all advantages are tem-
porary and uncertain because firms interact and the
“perennial gale of creative destruction” erodes past
accomplishments (1976: 89). Successful action
evokes reaction from rivals. Indeed, it is the dy-
namic process of firm actions and rival actions that
defines the market process. Firms are spurred to
engage in a cycle of action as they continually seek
to learn more about action-performance relation-
ships. When a firm leads with a new product or
service, it puts pressure on competitors’ products
and services, perhaps to the point of rendering
them obsolete. Those competitors must act if they
are to stay viable. Some rivals imitate, and others
make innovative thrusts of their own. Regardless,
the cycle repeats again and again as rivals struggle
for profits and market share. It is this competitive
interaction of rivals in pursuit of profits that results
in market progress and evolution (Schumpeter,
1976).

Recent research supports the coevolutionary na-
ture of action and reaction. Specifically, in a variety
of different studies, conducted in a variety of dif-
ferent industries, researchers have found a positive
correlation between firm actions and rival actions.
Indeed, response time, often a measure of compet-
itive intensity and rival action speed, has ranged
from as low as 8 days in the airline industry to 2

4

days in computer retailing and 124 days in high-
technology industries (Grimm & Smith, 1997).

Thus, Red Queen theory recognizes the interde-
pendent nature of firms. Specifically, the search,
action, and learning process does not end with a
focal firm’s actions (Barnett & McKendrick, 2004).

That is, the improved performance of the firm may
come at the expense of rivals’ performance, which,
in turn, may prompt rivals to search, act, and learn
to improve their own performance. In this sense,
learning and competition are codependent. To-
gether, they explain the incremental and relative
process by which firms evolve as they try to im-
prove performance.

Thus, we predict:

Hypothesis 1b. As the number of focal firm
actions increases, the number of rival firm ac-
tions and the speed of rival actions increase.

A number of studies have focused on the perfor-
mance consequences of industry rivalry. In a sam-
ple of software firms, Young and colleagues (1996)
found that as industry rivalry, measured as number
of rival actions, increased, focal firm performance
decreased. Chen and Miller (1994) found that
higher levels of rival responses decreased perfor-
mance in the airline industry, and Schomburg,
Grimm, and Smith (1994) found a negative relation-
ship between rivalry and profitability in the beer,
telecommunications, and personal computer in-
dustries. Finally, Smith and colleagues (1992)
found that increased competitive actions were re-
lated to lower profitability in the airline industry.

Evolutionary scholars have also examined the
performance consequences of Red Queen competi-
tion. Barnett and Hansen (1996) argued that a focal
firm’s superior performance leads a rival to search
for new opportunities to improve its own perfor-
mance. Assuming effective actions are found, the
rival’s position improves at the expense of the focal
firm. In a study of Illinois banks from 1900 to 1992,
Barnett and Hansen (1996) found that a focal firm’s
own competitive experience increased its chances
of success and survival, whereas its rivals’ aggre-
gate relative experience decreased the focal firm’s
success. They argued that firms are constrained by
their history, falling into competency traps where
they respond to new developments with old ac-
tions (Ingram, 2002; Levitt & March, 1988).

2

In view of this theory and research, we
hypothesize:

Hypothesis 1c. With the number of focal firm
actions held constant, as the number of rival
firm actions and the speed of rival actions in-
crease, focal firm performance decreases.

2 As noted, although the long-term effects of competi-
tion on performance might be positive (Porter, 1980), we
believe the shorter-term effects will be negative.

64 FebruaryAcademy of Management Journal

Hypotheses 1a–1c represent our baseline Red
Queen prediction on the positive and negative con-
sequences of firm and rival actions for focal firm
performance. We next consider how this Red
Queen relationship may be affected by industry
conditions and market position. Figure 1 is an il-
lustration of the hypothesized relationships.

MODERATORS OF THE RED QUEEN EFFECT

We draw from evolutionary theory (Nelson &
Winter, 1982), the industry position school of com-
petitive advantage (Porter, 1980), and action re-
search (Smith et al., 1992) to propose how industry
concentration, industry demand conditions, and
market position moderate the baseline model re-
garding the relationship between focal firm action,
rival action, rival action speed, and firm perfor-
mance. We theorize that these factors moderate the
relationship between focal firm action, rival action,
rival action speed, and focal firm performance by
affecting the ability of a focal firm and rival firms to
learn from search and action.

Industry Concentration

Industry concentration, commonly measured by
the percentage of the market share held by the
largest firms in an industry (the Herfindahl index),
is an important industry characteristic. As Wald-
man and Jensen noted, “Seller concentration
within a particular market is regarded as a signifi-
cant aspect of market structure because of its hy-
pothesized relationship to market power and, ulti-
mately, to behavior and performance” (2001: 94).
Theory from economics suggests that a small num-
ber of dominant firms in an industry will recognize
their mutual dependence and tacitly coordinate
search and action in an effort to limit competition

and rivalry (Scherer & Ross, 1990). The implicit
motive for this coordination is that escalation of
competition increases costs and hurts performance.
Conversely, as the number of firms increases,
search, action, and potential learning increase as it
becomes more and more difficult for this coordina-
tion to occur (Williamson, 1965). Under these con-
ditions, it is more challenging for a focal firm to
find unique opportunities to act and, as a conse-
quence, effective search, action, and learning be-
come more costly. One implication of this argu-
ment is that the effects of focal firm search, action,
and learning on performance are greater in more
concentrated industries. In such environments,
search and action are less frequent, and so it is
easier for firms to learn and to comprehend the
consequences of their actions as they receive more
attention from market participants. Moreover,
when a market consists of just a few large firms,
customers have limited choices (fewer competi-
tors), and they are therefore more likely to be at-
tracted to the new actions of dominant firms.

For the same reasons that a firm’s performance
gains from action are likely to be greater in concen-
trated industries, because of the high market shares
of firms, limited choices of customers, and ease of
learning from the effectiveness of search and ac-
tion, the effects of rivals’ actions on the focal firm’s
performance are also greater. Firm actions in highly
concentrated industries are much more likely to
garner the attention of competitors because mutual
awareness is very high (Bain, 1951). Thus, in highly
concentrated industries, rival firms are more apt to
learn of the actions of a focal firm and to respond to
those actions to stave off the negative consequences
of nonresponse. In addition, and perhaps more im-
portantly, in highly concentrated industries, rival
firms are more inclined to respond, and to respond
quickly, to teach their competitors that breaking the

FIGURE 1
Hypothesized Relationships

2008 65Derfus, Maggitti, Grimm, and Smith

unwritten covenant of tacit collusion will be pun-
ished severely. On the other hand, in less concen-
trated industries, where there are more competitors
to keep track of, actions are less likely to provoke
responses because rivals will not be aware of
the initial behavior. As Scherer and Ross stated,
“As the number of sellers increases and the share
of industry output supplied by a representative
firm decreases, individual producers are increas-
ingly apt to ignore the effect of their price and
output decisions on rival reactions” (1990: 277).
Thus, in concentrated industries, focal firm actions
will have a greater impact on performance, evoke
a larger number and higher speed of rival ac-
tions, and those rival actions and their speed will
have a greater impact on the focal firm’s per-
formance.

From an evolutionary perspective, Barnett and
Hansen (1996) contended that the Red Queen effect
would be less constraining when a firm faced a
relatively small number of competitively different
cohorts, as in the condition of high concentration.
These authors described how increases in the num-
ber of competitive relationships (i.e., lower concen-
tration) constrained effective learning and adapta-
tion. They noted that each constraint lowers the
likelihood that a focal firm can carry out effective
search, action, and learning, arguing that costs of
search and action come to outweigh the benefits.
Using a sample of Illinois banks, they found that
failure rates increased with the number of compet-
itive relationships. These findings suggest that de-
creased concentration increases the number of
competitive relationships to manage and that under
low concentration, a firm’s action has less impact
on both its own performance and rival action,
while rival action has less impact on focal firm
performance. Carroll and Hannan’s (1989) density
dependence theory also supports this prediction.
Specifically, this research showed that firms
founded under conditions of many competitors are
less likely to survive in the long term because of
scarcity of resources (Carroll & Hannan, 2000).

Given the above arguments, we propose:

Hypothesis 2a. Industry concentration posi-
tively moderates the relationship between a
focal firm’s actions and its performance.

Hypothesis 2b. Industry concentration posi-
tively moderates the relationship between a
focal firm’s actions and rival actions and the
speed of rival actions.

Hypothesis 2c. Industry concentration nega-
tively moderates the relationship between rival

actions, the speed of rival actions, and a focal
firm’s performance.

Industry Demand

The growth rate of industry demand should also
have an impact on Red Queen competition. Studies
by Caves (1980) and Bothwell, Cooley, and Hall
(1984) showed that firms in high-growth industries
are less concerned about competing with rivals be-
cause they are able to enhance revenues simply by
maintaining their shares of the steadily increasing
demand. Therefore, high industry growth leads to a
“live-and-let-live” attitude among firms (Bradburd
& Caves, 1982; Liebowitz, 1982). A growing market
facilitates existing routines, and each firm can in-
crease its share of the pie by searching for and
carrying out actions that it knows will work with-
out affecting rivals. Conversely, a decline in indus-
try demand will prompt firms to search for new
ways of generating demand, by instituting a new
price cut or new marketing campaign, that initiate
or escalate warfare (Caves, 1980).

Research exploring the evolution of industries
has examined the differing effects that the early,
high-demand, stage and the mature, decreasing de-
mand, stage of the industry life cycle have on com-
petition among firms (Agarwal & Gort, 1996; Agar-
wal, Sarkar, & Echambadi, 2002; Carroll & Hannan,
1989). Specifically this work speculates that during
periods of high demand growth firms take actions
that help create that demand and thereby, benefit
all the firms in their industry (Agarwal & Bayus,
2002). Carroll and Hannon (1989) described such
early-stage actions as “legitimizing actions,” as op-
posed to later-stage “competitive” actions.

In regard to Red Queen competition, high-growth
environments provide fertile ground for searching
and learning about new opportunities to act, and
limit the negative effect such actions have on com-
petitors. Increasing industry growth mitigates the
Red Queen argument that a firm’s performance
gains come at the expense of other firms (Barnett &
Hansen, 1996). For example, irrespective of rival
actions, a focal firm’s successful new-product
launch is going to be even more successful when
the number of consumers seeking such products is
growing. Accordingly, all the firms in a high-
growth industry will be focused on developing suc-
cessful “initial” actions and less focused on re-
sponding more often or faster to other firms’
actions. Further, when rivals do act, their actions
are more likely to increase industry growth overall
rather than have a deleterious impact on another
firm’s performance. Therefore, we propose:

66 FebruaryAcademy of Management Journal

Hypothesis 3a. Industry demand positively
moderates the relationship between a focal
firm’s actions and its performance.

Hypothesis 3b. Industry demand negatively
moderates the relationship between a focal
firm’s actions and rival actions and the speed
of rival actions.

Hypothesis 3c. Industry demand positively
moderates the relationship between rival ac-
tions, speed of rival actions, and focal firm
performance.

Market Position

We next theorize about the effect of the relative
market position of a focal firm. Research on the Red
Queen effect has suggested that market leaders are
less affected by Red Queen competition than are
other firms. For example, Barnett and McKendrick
(2004) found that market share leaders—in their
case, large firms—were the most likely to act to
develop new products in the disk drive industry.
They also found that, when exposed to compe-
tition, large firms were less likely to fail. How-
ever, they also noted that market share leaders
can become isolated from competition and, when
this happens, their survival may be threatened as
they lose their ability to learn from search and
action.

Action research also suggests that market leaders
more effectively search and act than their rivals and
react more quickly than their rivals (Smith et al.,
2001). Presumably, market leaders have the re-
sources to engage in more effective search and ac-
tion, which facilitates greater learning. In essence,
this is how they obtain and defend their market
positions. Ferrier, Smith, and Grimm (1999) found
that persistent market leaders act more frequently,
faster, and with more complexity. The actions of
market leaders should be more positively related to
performance than the actions of other firms in an
industry because they have more experience and
enjoy more efficient search and action routines. In
a sense, they have more effectively institutional-
ized the search, action, and learning process. Spe-
cifically, actions of market leaders are more visible
to customers and therefore likely to garner more
customer attention (Smith et al., 1992). Under these
conditions, managers can more effectively learn
from their actions. Young et al. (1996) found that
market leaders benefit from significant scale effects
of action that smaller firms cannot obtain. For mar-
ket-leading firms, the cost of search, action, and
learning can be spread over a larger customer base.

Although different predictions are possible re-

garding how rivals will behave with regard to mar-
ket leaders,3 we believe the more powerful argu-
ment is that rivals are unlikely to act or act quickly
against leading firms because of fear of retribution
(Scherer & Ross, 1990). Specifically, research in
industrial-organization (IO) economics has investi-
gated the behavior of dominant firms with regard to
their rivals on several fronts; it has been found that
pricing (Gaskins, 1971; Kamien & Schwartz, 1971),
R&D and patenting (Gilbert & Newberry, 1982),
product proliferation (Schmalensee, 1976), adver-
tising (Comanor & Wilson, 1967; Cubbin &
Domberger, 1988), and capacity increase (Masson &
Shannan, 1986; Spence, 1977) actions by dominant
firms deter rival entry. Ferrier and colleagues
(1999) found that market leaders that engaged in
more frequent, speedier actions and utilized more
complex action repertoires deterred the actions of
challengers. Thus, we expect that rivals will be
deterred from attacking an industry leader.

Finally, we predict that the frequency and speed
of rival actions, if they occur, have less negative
impact on leaders’ performance than on nonlead-
ers’ performance. Customers of leaders are more
likely to remain loyal in the face of rival actions.
Leader firms have stronger brand reputation and
customers are less likely to defect because the
switching costs of moving from a market leader are
potentially higher (Scherer & Ross, 1990). In sum-
mary, prior work shows that market leaders are
more effective than nonleaders, and their actions
have a stronger impact on performance than the
actions of nonleaders. Because leaders’ actions are
more likely to deter, rather than provoke rivals,
they evoke fewer and slower rival actions than do
nonleaders’ actions. And, because of customer loy-
alty and higher switching costs, rivals’ actions and
their speed do not detract from leader performance
as much as they do nonleaders’ performance.
Therefore, we predict:

Hypothesis 4a. Market position positively mod-
erates the relationship between a focal firm’s
actions and its performance.

Hypothesis 4b. Market position negatively
moderates the relationship between a focal
firm’s actions and rival actions and the speed
of rival actions.

Hypothesis 4c. Market position positively mod-
erates the relationship between rival actions,

3 One possible alternative explanation is that rivals are
more likely to follow or imitate market leader actions
because they are perceived as more legitimate.

2008 67Derfus, Maggitti, Grimm, and Smith

the speed of rival actions, and a focal firm’s
performance.

RESEARCH METHODS

Sample

To test the hypotheses, we developed a sample of
all the major competitors in 11 different industries
across a broad spectrum of the U.S. economy. One
requirement for sample inclusion was that firms
were competing in the same markets so that their
specific actions and firm performance could be di-
rectly connected to the competition and perfor-
mance in these markets. As a result, we focused
solely on the actions and performance of U.S. firms
and included only those industries in which 7

0

percent or more of industry sales was generated by
firms that were public, had a distinct single-busi-
ness entity competing in the specific U.S. market,
and reported performance relative to that U.S. mar-
ket. Firms needed to meet these criteria so that we
could match the actions of their single-business
entities with their performance in only in
that market/industry.

Eleven industries met our criteria, including ap-
pliance manufacturing, athletic footwear manufac-
turing, automobile manufacturing, brewing, gen-
eral retailing, book retailing, lumber and hardware
retailing, long-distance telephone services, steel
manufacturing, and grocery retailing. With repre-
sentation from manufacturing, services, and retail-
ing, good industry variation was achieved. On
average, included firms accounted for 87 percent
of U.S. industry sales in their respective market/
industry.

Data Collection: Competitive

Actions

Competitive actions are defined as specific and
observable moves, such as new marketing cam-
paigns or new-product introductions, initiated by a
firm to defend or improve its relative competitive
position (Chen, 1988; Smith et al., 1992; Young et
al., 1996). Actions that are observable to customers,
competitors, and other industry watchers are most
likely to be reported in the business press (Miller &
Chen, 1994) and thereby are available for identifi-
cation, data collection, and analysis. We identified
and coded observable competitive actions by con-
ducting a structured content analysis (Jauch, Os-
born, & Martin, 1980) of newspaper and trade mag-
azine articles found on the Lexus-Nexus article
index. This index allows electronic searching of
full-text articles from thousands of newspapers and
journals. For each of the industries chosen, at least

one industry trade magazine was searched. Addi-
tionally, the New York Times and the Wall Street
Journal were searched for all industries. We iden-
tified 76,963 article citations via keyword search-
ing of the Lexus-Nexus database. Coders then con-
tent-analyzed the full texts of articles that
potentially contained reports of competitive ac-
tions. Only the earliest report of an action was
entered into the database. This procedure resulted
in a database containing 4,474 actions. To verify
the accuracy of the coding, two coders reviewed 10
percent of the article citations for each industry.
Action identification and action-type coding agree-
ment were obtained for 99.25 percent of the over
7,697 citations they read.

Actions were collected for 58 firms; missing data
reduced the sample to 56 firms over a six-year
period, 1993 through 1998. The mean number of
actions per firm was 12.58. The maximum number
of actions per firm per year was 51; the minimum,
0. The most common actions related to pricing, and
the least common were geographic actions.

Firm financial data, firm size, and industry con-
text variables were collected from Standard &
Poor’s Compustat database, which offers financial
data on all companies that were publicly traded on
North American stock markets in those years.
These data are collected from annual reports, Secu-
rities and Exchange Commission (SEC) filings, and
other publicly available documents. Where neces-
sary, we adjusted financial data to remove contri-
butions from non-U.S. operations, thereby match-
ing the financial figures to the actions accounted
for in the study. These adjustments were possible
because Compustat offers detailed geographic seg-
ment data delineating company operations in vari-
ous countries.

Measures

Focal firm total actions and rival total actions.
Five types of focal firm and rival actions were mea-
sured: pricing, capacity, geographic, marketing,
and product introductions. We calculated firm total
actions or activity by simply summing the number
of all five actions for a focal firm in a given year. We
then operationalized rival total actions or compet-
itive activity by subtracting a focal firm’s total num-
ber of actions in a given year from the total number
of actions taken by all competitors in a
focal industry.

Rival action speed. As in other competitive dy-
namics research (e.g., Ferrier et al., 1999; Young et
al., 1996), rival firm action speed quantified the
average length of time it took rivals to act after a
focal firm acted. To calculate this measure, we de-

68 FebruaryAcademy of Management Journal

termined the number of days between each firm
action and the first rival action and then averaged
those scores for each focal firm for each year. Fi-
nally, we took the reciprocal of this value to aid in
interpretation of results. The resulting measure
equates high rival action speed values to fast rival
action speed and low rival action speed values to
slow rival action speed.

Focal firm performance. Focal firm perfor-
mance was operationalized with accounting mea-
sures of return on sales (ROS) and return on assets
(ROA) in the same year as the action measure.

Industry conditions. Industry concentration and
industry demand were used to capture the industry
context in which firm and rival actions took place.
These measures served as independent or control
variables in all regressions and were also interacted
with firm actions, rival actions, and rival action
speed in tests of Hypotheses 2a–2c and 3a–3c. In-
dustry concentration was calculated as the Herfin-
dahl measure of the market shares of the firms in
each industry for each year. Industry demand was
measured as industry growth, defined as the per-
cent change in sales from the previous year to a
focal year.

Relative market position. Relative market posi-
tion was measured as rank order based on market
share for each firm in each industry for each year.
This variable was used as an independent or a
control variable in all regressions and was also
interacted with firm actions, rival actions, and rival
action speed in testing Hypotheses 4a– 4c.

Control variables. To control for unobserved dif-
ferences in industry factors that might influence
market dynamics, we included industry dummies
in all regressions.

In this study, we had two basic regression mod-
els. In the first, we regressed our independent ac-
tion variables on firm performance. However, since
we had two measures of firm performance, return
on assets and return on sales, a model is presented
for each. In these models, we lagged a focal firm’s
prior year return on assets or return on sales per-
formance, respectively, to control for the influence
the variable might have on performance in the fol-
lowing year and also to help control for correlated
error terms of our longitudinal data (Young et al.,
1996). In these regressions, we also controlled for
firm characteristics that have been shown to influ-
ence firm actions and performance, including size
and slack resources (Smith et al., 1992). Specifi-
cally, sales measured size, and the quick ratio mea-
sured slack resources (e.g., Ferrier et al., 1999). The
logic was that firms with more assets and resources
are able to undertake more actions (Smith et al.,

2001). We then repeated these regressions replac-
ing rival action speed with rival total actions.

When we tested the effect of focal firm actions on
rival actions and rival action speed, we controlled
for prior year aggregate performance utilizing
lagged rival firm prior year return on sales. This
lagged composite was calculated as the aggregated
net income of all rival firms in an industry divided
by their aggregated sales. As in our regressions on
firm performance, we also controlled for rival size
and rival slack resources. Rival’s size was calcu-
lated as an aggregate average measure of the relative
size of each firm’s pool of rivals. It is an annual sum
of the sales of the unique group of rivals pertaining
to each focal firm. Rival quick ratio is a composite
average of the quick ratios of those rival firms.

We followed the practice of prior competitive
dynamics researchers by investigating the impact
of actions on the same year’s performance (e.g.,
Ferrier et al., 1999; Young et al., 1996). During the
years of our study, for these 11 industries, the av-
erage number of days between a focal firm action
and a rival action is 12.3 (s.d. � 19.3) and the
maximum number of days between actions is 149.
This relatively short time frame provided further
support for our same-year analysis of actions and
performance.

In calculating the measures in this study, it was
important to precisely define the focal and rival
firms. For example, in the U.S. brewing industry
there were three sampled firms: Anheuser-Busch,
Miller Brewing, and Adolph Coors. When An-
heuser-Busch was the focal firm, Miller Brewing
and Adolph Coors were the rival firms. Similarly,
when Miller Brewing was focal, Anheuser-Busch
and Adolph Coors were the rivals. We calculated
firm total actions and performance measures for
each firm individually and the rival total actions
measure for each firm’s unique set of rivals. Since
the unit of observation was the firm-year, changes
in the rival set from year to year must be accounted
for. Therefore, the rival set was officially defined as
all the other firms competing in the focal firm’s
industry for the year under consideration.

RESULTS

Table 1 reports the means and correlations
among all variables in this study. We tested hy-
potheses with random-effects regression models to
ensure that error due to serial correlation in our
panel data set was specified and analyzed (Erez,
Bloom, & Wells, 1996). In addition, we used nega-
tive binomial regression analysis for regression
models in which the dependent variable was rival
total actions, a count-type variable. This type of

2008 69Derfus, Maggitti, Grimm, and Smith

T
A

B
L

E
1

D
es

cr
ip

ti
v
e

S
ta

ti
st

ic
sa

V
a
ri

a
b
le

b
M

e

a
n

s.
d

.
1

2

3

4
5

6

7

8

9

1
0

1
1

1
2

1
3

1
4

1
5

1
6

1
7

1
8

1
9

2
0

2
1

2
2

2
3

2
4

2
5

2
6

2
7

2
8

2
9

3
0

3
1

3
2

3
3

1
.

R
O

A
3
.7

4

5
.5

1
2
.

R
O

S
4
.5

3

5
.3

3
.4

5
3
.

F
ir

m
to

ta
l

ac
ti

o
n

s
1

3
.1

8
1

0
.6

5
.2

2
.1

4
4
.

R
iv

al
to

ta
l
ac
ti
o
n

s
6

3
.5

7
4
4
.6

9
�

.1
3

�
.3

9
.3

1

5
.

R
iv

al
ac

ti
o
n

sp
ee

d

0
.1

8
0
.1

4
�

.1
1

�
.3

2
.4

6
.9

5
6
.

F
ir

m
m

ar

k
et

sh
ar

e
ra

n
k

0
.3

8

0
.2

7
.0

4
�

.0
7

.1
5

.3
4

.3
2

7
.

I

n
d

u
st

ry
H

er
fi

n
d

ah
l

0
.2

2
0
.1

8
.0

8
.3

8
.0
9
�

.5
4

�
.4

9
�

.5
5

8
.

In
d

u
st

ry
gr

o
w

th

0
.0

9
0

.1

1
.0

3
�

.1
5

�
.1

5
�

.1
9

�
.2

2
�

.0
8

.1
4

9
.

M
ar

k
et
sh
ar
e
ra
n
k

�
fi

rm
ac

ti
o
n

s
5
.5

5

6
.1

9
.1

9
.0

1
.7

2
.3

5
.4

2
.6

7
�

.2
8

�
.0

8

1
0
.

M
ar
k
et
sh
ar
e
ra
n
k

�
ri

v
al

ac
ti
o
n

s
2

8
.6

6
3

2
.2

0
�

.0
7
�
.2
5
.2
0
.6

9
.6

3
.8

5
�

.6
2

�
.0
9
.6
4

1
1
.

H
er

fi
n

d
ah

l
�

fi
rm

ac
ti
o
n

s
3
.4

7

4
.4

2
.1

6
.4

0
.7

0
�
.1
5
�
.0

1
�

.2
1

.6
2
�
.1

1
.2

1
�

.2
6

1
2
.
H
er
fi
n
d
ah
l
�

ri
v
al

ac
ti
o
n

s
1

1
.9

4

1
0
.0

1
�
.0
7

.

0
2

.3
8

.3
5

.4
1

�
.3
4
.4

6
�

.1
3
�
.0
4
�
.1
3

.4
6

1
3
.

In
d
u
st
ry
gr
o
w

th
�

fi
rm
ac
ti
o
n

s
1
.0

9
1
.5

5
.2
0
�

.0
6

.3
8
.0
8
.1
3

.1
2

.0
0

.5
4

.3
6

.1
1
.1
3
.0
0

1
4
.

In
d
u
st
ry
gr
o
w
th
�
ri
v
al
ac
ti
o
n

s
5
.2

2
5
.7

9
�
.0
8
�
.3
3
.1

1
.6

0
.5

3
.2

6
�
.3
2

.4
2

.2
3

.5
0

�
.2
0
.1
3
.5
3
1
5
.
M
ar
k
et
sh
ar
e
ra
n
k
�
ri
v
al
sp
ee

d
0
.0

8
0
.0

9
�
.0
6
�
.2
2
.2
8
.6

7
.6

6
.8

7
�

.6
1

�
.1

2
.7

1
.9

8
�

.2
1
�
.1
0
.1
3
.4
6

1
6
.

H
er
fi
n
d
ah
l
�
ri
v
al
sp
ee
d
0
.0

4
0
.0

3
�

.0
5

.1
0

.5
6

.3
4

.4
8

�
.2

8
.4

1
�

.1
6

.0
7
�
.1
1
.6

3
.9

4
.0

6
.1
1
�
.0
5

1
7
.

In
d
u
st
ry
gr
o
w
th
�
ri
v
al
sp
ee
d
0
.0
1
0
.0
2
�
.0
6
�
.3
0
.2

2
.5

9
.5

8
.2

4
�

.2
9

.4
0

.2
8
.4
6
�
.1
2
.1
8
.6

4
.9

7
.4

5
.2
0

1
8
.

L
ag

ge
d

R
O

A
0
.0

4
0
.0

6
.5

2
.2

2
.0

6
�
.0
5
�
.0
7
.0
0
.0
3
�

.0
2

.0
7
�
.0
6
�
.0
2
�
.0
1
.0
8
�

.0
9

�
.0
7
�
.0
5
�
.0

9
1
9
.

L
ag
ge
d
R
O

S
0
.0

4
0
.0
5
.3
1
.6
4
.0
1
�

.3
3

�
.2
9
�
.0
9
.2
6
�
.1
7
�
.0
7
�
.2

4
.1

8
.0
1
�
.0
9
�
.3
5
�

.2
2

.0
4

�
.3
5
.5

4
2
0
.

L
ag
ge
d
ri
v
al
R
O
S
0
.0

5
0
.0

4
.0
5
�
.1
6
�
.0
5
�
.0
4
�
.1
5
�
.0
8
.0
5
.0
9
�
.0
4
�
.0
8
�
.0
6
�
.0
6
�
.0
1
�
.0
9
.1
3
�
.1
5
�
.0
9
.1
3

.0
3

2
1
.

F
ir

m
sa

le
s

1
5
,6

6
3

2
5
,1

9
0
.1
5
.1
0
.4
6
.0
5
.1
4
.0
2
.1
1
�
.1
1
.2
8
.0
0
.3
9
.1
6
.1
4
�
.0
4
.0
5
.2
9
�
.0
4
.0

6
.0

3
�
.1
0

2
2
.

F
ir

m
q
u

ic
k

ra
ti

o
0
.6

0

0
.4

3
.1

1
.3

0
�
.0
7
�
.1
4
�
.0
8
�
.0
6
.1
0
�
.1
2
�
.1
5
�
.1
1
.1
4
.0
9
�
.0
6
�
.1
0
�
.0
9
.1
1
�
.1
0
.1
1
.2
1
�
.2
5
�
.1
6

2
3
.

R
iv

al
sa

le
s

8
2
,2

0
2

9
1
,6

3
0
�
.0
9
�
.2
2

.3
1

.6
1

.6
6

�
.0
5
�
.2
6
�
.2
0
.1

4
.2

0
�
.0
4

.3
9

.0
6

.2
4

.2
2

.4
4

.2
4
�
.0
8
�
.1
9
�
.0
3
.4
3
�
.2
2

2
4
.

In
d
u
st

ry
q
u

ic
k
ra
ti
o
0
.6

1
0
.3

1
.0

3
.3

2
�
.0
6
�
.1
5
�
.0
8
.0
2
.0
8
�
.1
9
�
.1
0
�
.0
6
.1

5
.1

0
�
.1
2
�
.1
2
�
.0
3
.1
2
�
.1
2
.0
0
.2
1
�
.3
4
�
.1

4
.7

4
.2

9
2
5
.

In
d
u
st

ry
1

0
.0

6
0
.2

4
.0

9
.2

7
.0
0
�
.2
3
�
.2
0
�
.1
6
.2
9
�
.1
8
�
.1
0
�
.1

8
.1

5
�
.0
3
�
.1
2
�
.1
9
�
.1
8
.0
0
�
.1
9
.0
6
.2
1
�
.2
0
�
.0
7
�
.0
1
�

.1
8

�
.0

1
2
6
.

In
d
u
st

ry
2

0
.0

7
0
.2

6
.2

8
.0

7
.1

4
�
.0
6
�
.0
8
�
.1
2
.2
4
.0
3
.0
0
�
.1
3
.2

4
.3

1
.1

7
.0
1
�
.1
4
.2
2

.0
1

.2
6
.1
1
.1
2
�
.1
5
.3
6
�
.2
3

.4
5

�
.0

6
2
7
.

In
d
u
st

ry
3

0
.1

9
0
.3

9
�

.2
0

�
.1
5
�
.2
1
.2
1
.2
1

.5
8

�
.5

3
�
.1
4
.1
4

.4
7

�
.3
2
�
.3
5
�
.1
0
.1
7
.4
9
�
.3
1

.1
7

�
.1
4
�
.1
2
�
.2
9
�
.2
6
.2
8
�
.2
2
.4
1
�
.1
2
�
.1

4
2
8
.

In
d
u
st

ry
4

0
.0
6
0
.2
4
�
.0
4
.2
3
�
.2
4
�
.3
2
�
.2
9
�
.1
9
�
.0
8
�
.1
3
�
.2
0
�
.2
1
�
.1
5
�

.2
5

�
.1
5
�
.2
1
�
.2
2
�
.2
2
�
.2
1
�
.0
6
.1
7
�
.2
0
�
.1
1
.2
2
�
.1
9
.3
1
�
.0
6
�
.0
7
.1

3
2
9
.

In
d
u
st

ry
5

0
.1

2
0
.3

2
�
.1
3
�
.1
2
.3
0
.4
2
.5
8
�
.1
5
�
.0
1
�
.0
5
.0
3
.0
1
.1
5

.5
9

.1
6
.2
6
.0
7

.6
8

.2
6
�
.1
1
.1
2
�

.3
0

.2
9
.0
0

.7
2

.0
0
�
.0
7
�
.0
8
�
.1
5
�
.0
7

3
0
.

In
d
u
st

ry
6

0
.0
6
0
.2
3
�
.0
1
.4
0
.1
8
�
.1
7
�
.1
3
�
.1
9
.6
2
�
.1
3
�
.0
5
�
.1
9

.5
3

.3
5
�
.1
6
�
.2
4
�
.1
8
.3
5
�
.2
4
�
.0
6
.2

5
.0

3
.1
8
.1
2
�
.0
9
.1
6
�
.0
6
�
.0
7
�
.1
4
�
.0
7
�
.0

8
3
1
.

In
d
u
st

ry
7

0
.0

8
0
.2

8
.0
3
�
.0
2
�
.1
4
�
.2
4
�
.2
6
�
.0
1
.0
7
.1
6
�
.0
9
�
.1
8
�
.0
6
�

.1
.0

4
�
.0
3
�
.1
9
�
.1
2
�
.0

3
.0

3
�
.0
1
.2
9
�
.0
8
�
.2
2
�
.1
9
�
.3
5
�
.0
7
�
.0
8
�
.1
6
�
.0
8
�
.0

9
3
2
.

In
d
u
st

ry
8

0
.0
8
0
.2
7
.0
2
�
.0
3
.0
1
�
.1
7
�
.1
8
�
.1
5
.0
2
�
.0
5
�
.0
7
�
.1
8
.0
1
�
.0
4
.0
0
�
.1
1
�
.1
9
�
.0
6
�
.1
1
�
.0
1
�
.0
3
�
.0
2
.5
4
�
.0
6
.5
1
�
.0
8
�
.0
7
�
.0
8
�
.1
5
�
.0
7
�
.0
8
�
.0
8
�
.0

9
3
3
.

In
d
u
st

ry
9

0
.1

5
0
.3

6
.1
3
�
.2
1
.2
1
.5
6
.4
4
.2
2
�
.4
3
�
.1
4
.3
3
.4
5
�
.2
1
�
.1
4
.0
0
.1
7
.4
0
�
.1
6
.1
7
.1
2
�
.1
7
.3
5
.0
2
�
.4
1
.3
5
�

.5
2

�
.1
1
�
.1
2
�
.2
4
�
.1
1
�
.1
3
�
.1
3
�
.1
4
�
.1
3

3
4
.

In
d
u
st

ry
1
0

0
.0
7
0
.2
6
�
.1
1
�
.1
3
�
.0
5
�
.1
5
�
.1
8
�
.1
5
.2
2
.1
4
�
.0
8
�
.1
7
.0
6
.1
1
.0
8
.0
3
�
.1
8
.0
2
.0
3
�
.0
7
�
.0
4
.0
8
�
.1
6
�
.1
4
�
.2
2
�
.2
0
�
.0
6
�
.0
7
�
.1
3
�
.0
6
�
.0
7
�
.0
7
�
.0
8
�
.0
7
�
.1
2
a
n

�
2
8
1
.

C
o
rr

el
at

io
n

s
ab

o
v
e

.1
1

ar
e

si
gn

if
ic

an
t

at
p

�
.0
5
.

b
R

O
A

an
d

R
O

S
ar

e
p

er
ce

n
ta

ge
s.

S
al

es
ar

e
in

m
il

li
o
n

s
o
f

d
o
ll

ar
s.

regression was used in these models for two rea-
sons. First, these count data are not normally dis-
tributed, violating a key assumption of generalized
least squares (GLS) regression analysis (Greene,
1993). Secondly, as is typically the case, our count
data are overdispersed, meaning the variance of the
event counts exceeds their means (Cameron & Tra-
vendi, 1986). A likelihood-ratio test of overdisper-
sion also indicated that negative binomial regres-
sion was an appropriate choice. Negative binomial
regression overcomes distribution problems and es-
timates an additional parameter that corrects for
overdispersion (Frome, Kutner, & Beauchamp,
1973). Tables 2 and 3 report the regression results.

Table 2 reports the results for regressions relating
firm actions, firm performance, rival actions, and
speed of rival actions. Table 3 reports the regres-
sion results that examine the impact of industry
environment and market leadership on the rela-
tionships between firm actions, firm performance,
rival actions, and speed of rival actions.

Hypothesis 1a states that as firm total action in-
creases, firm performance increases. This hypothe-
sis is fully supported. As seen in models 1 and 3 in
Table 2, firm actions have a positive, significant
coefficient for both return on assets (� � 0.08, p �
.01) and return on sales (� � 0.08, p � .01). These
results are repeated in models 2 and 4 of Table 2, in
which rival action speed replaces rival actions in
the models. That is, firm actions again have a pos-

itive, significant effect on both return on assets (� �
0.12, p � .01) and return on sales (� � 0.11, p �
.01).

Hypothesis 1b states that as firm actions increase,
rival actions and rival action speed also increase.
This hypothesis is supported. Specifically, models
5 and 6 in Table 2 report a significant, positive
coefficient for the relationship between firm ac-
tions and rival actions (� � 0.01, p � .05) and firm
actions and rival speed (� � 0.00, p � .01),
respectively.

Hypothesis 1c states that, as rival actions and
rival action speed increase, focal firm performance
decreases. This hypothesis is also fully supported.
As reported in Table 2, rival actions is significantly
and negatively related to focal firm ROA in model 1
(� � �0.05, p � .01) and ROS in model 3 (� �
�0.03, p � .01). Similarly, as shown in models 2
and 4, rival action speed is significantly and nega-
tively related to focal firm’s ROA in model 1 (� �
�17.02, p � .01) and ROS in model 2 (� � �12.93,
p � .01).

Hypotheses 2a–2c, 3a–3c, and 4a– 4c explore the
boundary conditions of the first hypothesis set, Hy-
potheses 1a–1c. Specifically, in the second and
third sets of hypotheses we examine how various
industry situations condition the relationship be-
tween firm actions, rival actions, rival action speed,
and firm performance. In the fourth set of hypoth-
eses, we look at how these relationships differ on

TABLE 2
Results of Random-Effects Regression Analyses of the Main Relationshipsa

Variables
Model 1:

ROA
Model 2:

ROA
Model 3:

ROS

Model 4:

ROS

Model 5:
Rival Total

Actions

Model 6: Rival
Speed of
Actions

Lagged ROA 35.74** (4.81) 35.73** (4.82)
Lagged ROS 38.78** (4.34) 39.93** (4.31)
Lagged rival ROS �2.14* (1.03) �0.30** (0.10)
Firm sales 0.29† (0.21) 0.30† (0.24) �0.14 (0.17) �0.14 (0.17)
Quick ratio 1.84* (0.75) 1.89* (0.75) 2.03** (0.60) 2.05** (0.60)
Rival sales �0.00* (0.00) �0.00* (0.00)
Industry quick ratio 0.22 (0.16) 0.03 (0.02)
Market share rank 0.93 (1.27) 0.84 (1.27) 2.95** (1.04) 2.81** (1.03) �0.13* (0.18) �0.04** (0.01)
Herfindahl index �5.52 (5.57) �5.78 (5.59) �1.19 (4.61) �1.17 (4.63) �0.34 (0.31) �0.12* (0.07)
Industry growth 6.12* (3.11) 5.06* (3.06) 2.83 (2.51) 2.29 (2.49) 1.19** (0.29) 0.06* (0.04)
Firm total actions 0.08** (0.03) 0.12** (0.04) 0.08** (0.03) 0.11** (0.03) 0.01* (0.00) 0.00** (0.00)
Rival total actions �0.05** (0.01) �0.03** (0.01)
Rival speed of actions �17.02** (5.21) �12.93** (4.23)

Constant 1.66 (2.54) 1.99 (2.55) 0.58 (2.13) 0.66 (2.14) 1.97** (0.58) 0.05 (0.04)
Wald chi-square 187.20** 185.18** 394.58** 406.50** 248.93** 2,117.46**

a Standard errors are in parentheses. Industry dummy variables were included in all regression models. Results are available upon
request. n � 281.

† p � .10
* p � .05

** p � .01

2008 71Derfus, Maggitti, Grimm, and Smith

the basis of the market share of a focal firm relative
to the other firms in the industry. Table 3 reports
the results of our tests of Hypotheses 2a–2c, 3a–3c,
and 4a– 4c.

Hypothesis 2a predicts industry concentration
positively moderates the relationship between fo-
cal firm actions and focal firm performance. This
hypothesis is partially supported. Specifically, and
as predicted, the effect of the interaction of indus-
try concentration on the relationship between focal
firm actions and firm performance is positive and
significantly related to firm ROS in models 3 and 4
of Table 3 (� � 0.59, p � .01; � � 0.50, p � .01).
There was no significant finding with respect to
focal firm ROA. This result suggests that the posi-

tive effect of focal firm actions on performance, in
terms of return on sales, is higher in concentrated
industries than in nonconcentrated industries.

In Hypothesis 2b, we predict that industry con-
centration positively moderates the relationship
between focal firm actions and both rival actions
and speed of rival action. Although there were no
significant findings with respect to rival actions,
the results shown in model 6 of Table 3 run counter
to this hypothesis for speed of rival action (� �
�0.01, p � .01). That is, in concentrated industries
the relationship between firm actions and speed of
rival action is weaker than it is in less concentrated
industries. No significant findings were found to
support or refute our Hypothesis 2c, that industry

TABLE 3
Results of Random-Effects Regression Analyses of Interactionsa

Variables
Model 1:
ROA
Model 2:
ROA
Model 3:

ROS
Model 4:

ROS
Model 5: Rival
Total Actions

Model 6: Rival
Speed of
Actions

Lagged ROA 34.48** (5.04) 34.12** (5.09)
Lagged ROS 42.02** (4.30) 42.23** (4.34)
Lagged rival ROS �2.18** (0.65) �0.30** (0.10)
Sales 0.24 (0.23) 0.25 (0.23) �0.31* (0.18) �0.33* (0.18)
Rival sales �0.00 (0.00) �0.00 (0.00)
Quick ratio 1.89* (0.76) 1.91* (0.77) 1.83** (0.59) 1.85** (0.59)
Industry quick ratio 0.06 (0.17) 0.02 (0.02)
Market share rank 5.40* (3.01) 5.53* (3.09) 5.78* (2.35) 5.75* (2.41) �0.42* (0.23) 0.01 (0.02)
Herfindahl index �4.89 (6.08) �5.42 (5.90) �5.11 (4.84) �3.90 (4.71) 1.53* (0.88) �0.05 (0.07)
Industry growth 0.36 (4.91) 1.20 (4.63) 2.34 (3.84) 2.97 (3.62) 3.61** (0.46) 0.14** (0.05)
Firm total actions 0.03 (0.11) 0.04 (0.12) �0.18* (0.09) �0.14† (0.09) 0.02** (0.01) 0.01** (0.00)
Rival total actions �0.02 (0.04) �0.04† (0.03)
Rival speed of actions �0.06 (13.65) �11.49 (10.65)
Herfindahl � firm total

actions
0.05 (0.18) 0.08 (0.18) 0.59** (0.14) 0.50** (0.14) 0.01 (0.02) �0.01** (0.00)

Herfindahl � rival total
actions

�0.04 (0.09) 0.10† (0.07)

Herfindahl � rival speed of
actions

�28.20 (36.17) 19.54 (28.36)

Industry growth � firm total
actions

0.11 (0.28) 0.15 (0.31) �0.11 (0.22) �0.17 (0.25) �0.14** (0.02) �0.01** (0.00)

Industry growth � rival total
actions

0.11 (0.10) 0.15* (0.08)

Industry growth � rival speed
of actions

13.43 (34.07) 34.13 (26.78)

Market share rank � firm
total actions

0.07 (0.14) 0.11 (0.15) 0.20* (0.11) 0.23* (0.12) �0.01* (0.01) �0.00** (0.00)

Market share rank � rival
total actions

�0.06* (0.03) �0.05* (0.03)

Market share rank � rival
speed of actions

�24.84* (13.5) �19.82* (10.44)

Constant 2.24 (3.01) 2.14 (3.00) 1.21 (2.40) 0.81 (2.39) �0.50 (0.68) �0.00 (0.047)
Wald chi-square 193.19** 189.51** 459.54** 453.06** 1,244.63** 2,242.03**

a Standard errors are in parentheses. Industry dummy variables were included in all regression models. Results are available upon
request. n � 281.
† p � .10
* p � .05
** p � .01

72 FebruaryAcademy of Management Journal

concentration negatively moderates the relation-
ship between rival firm actions and focal firm
performance.

Although we found no support for Hypothesis
3a, predicting that industry demand conditions
positively moderate the relationship between focal
firm actions and focal firm performance, our Hy-
pothesis 3b, in which we predict that industry de-
mand negatively moderates the relationship be-
tween focal firm actions and rival actions and rival
action speed, was supported. That is, models 5 and
6 of Table 3 indicate that the interaction between
focal firm actions and industry growth was nega-
tive and significantly related to both rival actions
(� � �0.14, p � .01) and the speed of rival actions
(� � �0.01, p � .01). Thus, and as predicted, as
industry demand increases, the effect of firm ac-
tions on rival actions and their speed declines.

Hypothesis 3c predicts that industry demand
positively moderates the relationship between fo-
cal firm performance and both rival actions and the
speed of rival actions. We found some support for
this hypothesis, as shown in model 3 of Table 3.
Specifically, the significant and positive effect of
the interaction between rival actions and industry
growth on focal firm return on sales (� � 0.15, p �
.05) is consistent with our hypothesis.

The influence that focal firm market position has
on rival actions and firm performance was explored
in Hypotheses 4a– 4c. In Hypothesis 4a, we predict
that market position positively moderates the rela-
tionship between focal firm actions and focal firm
performance. This hypothesis was partially sup-
ported, as indicated in models 3 and 4 of Table 3, in
which the interaction of market position and firm
actions is positive and significantly related to re-
turn on sales (� � 0.20, p � .05; � � 0.23, p � .05).
This result suggests that the positive impact of a
firm’s actions on performance is greater for firms
that have higher market shares in an industry.

Hypothesis 4b predicts that market position neg-
atively moderates the relationship between focal
firm actions and both rival actions and their speed.
This hypothesis is supported. That is, the interac-
tion of firm actions with higher market position is
negatively and significantly related to both rival
firm actions in model 5 of Table 3 (� � �0.01, p �
.05) and the speed of rival actions in model 6 of the
same table (� � �0.004, p � .01). Thus, the actions
of firms with higher market shares than their rivals
tend to not increase rival actions and rival action
speed as much as do the actions of firms with lower
market shares.

Hypothesis 4c predicts market position posi-
tively moderates the relationship between both ri-
val firm actions and rival firm action speed and

focal firm performance. Thus, we expected that
rival actions and rival action speed would not af-
fect market leaders in the same way as they would
affect the performance of nonleaders. Results did
not support this hypothesis and, in fact, were con-
trary to our expectation. Specifically, Table 3
shows that the interaction between market position
and rival actions is negatively and significantly
related to both focal firm return on assets in models
1 and 2 (� � �0.06, p � .05; � � �24.84, p � .05)
and focal firm return on sales in models 3 and 4
(� � �0.05, p � .05; � � �19.82, p � .05). Thus,
rival actions have a greater negative impact on
firms with larger shares of an industry market than
on firms with smaller market shares.

DISCUSSION

This study has shed light on Red Queen compe-
tition by investigating the relationships between
focal firm actions, rival firm actions, and focal firm
performance in a variety of industries. In line with
Red Queen theory, all the relationships in our base-
line model were supported and showed that even
though a focal firm’s actions do increase its perfor-
mance, they also increase the number and speed of
rivals’ actions which, at least partially, negatively
impact the focal firms’ performance. Indeed, to
paraphrase the Red Queen in Lewis Carroll’s
Through the Looking Glass (1960), it is true that the
firms studied here have “to run as fast as they can
to stay in place, and twice as fast as that” to get
ahead. Although portions of this baseline model
have been tested elsewhere, we are aware of no
previous study that has examined both the positive
and negative effects of actions, as we do in the
present study. Studying these effects together en-
ables us to elucidate the positive and negative as-
pects of action, and to clarify the relative importance
of these aspects with regard to firm performance.
Below we offer Figures 2 and 3 to illustrate this.

Plotting the data from the baseline models we
used to test Hypotheses 1a–1c, models 1– 4 in Table
2, we graphically present the Red Queen effect in
Figure 2. This graph illustrates the counter-balanc-
ing effects of firm and rival actions on firm perfor-
mance as measured by return on assets (ROA) and
return on sales (ROS). Specifically, we see the ac-
tual average number of firm actions and rival ac-
tions associated with various levels of perfor-
mance. As expected, performance gains from action
are maximized when firm actions are high and rival
actions are low. Perhaps less expected, the number
of rival actions necessary to seriously, negatively
impact firm performance is surprisingly high rela-

2008 73Derfus, Maggitti, Grimm, and Smith

tive to the number of firm actions necessary to
positively increase performance.

Figure 3 illustrates the strength of the Red Queen
effect by presenting the net incremental effect that
firm actions have on firm performance directly, and
indirectly, through rival actions and rival action
speed. The line with the steepest positive slope in
both the ROA and ROS plots represents the direct
effects of firm actions on performance. For ROA,
this slope is calculated as the mean of the coeffi-
cients for firm actions in models 1 and 2 of Table 1
(0.08 and 0.12, respectively).

The three lines below the “firm actions line” in
Figure 3 reveal the strength of the Red Queen effect

in our research. Specifically, line 2 shows how
much the direct positive effect of actions on perfor-
mance is reduced by the indirect negative impact
focal firm actions have when they stimulate rival
action. We calculated this negative impact by tak-
ing the derivative of rival total actions with respect
to firm total actions in model 5 multiplied by the
coefficient on rival total actions in model 1 (– 0.05).
Similarly, line 3 shows the direct positive effect of
actions on performance along with the indirect neg-
ative impact focal firm actions have by stimulating
rival speed of actions. This negative impact was
calculated as the derivative of rival speed of actions
with respect to firm total actions in model 6 mul-

FIGURE 2
Effects of Firm Actions and Rival Actions on Firm Performance

FIGURE 3
Incremental Effects of Firm Actions on Firm Performance

74 FebruaryAcademy of Management Journal

tiplied by the coefficient on rival speed of actions
in model 2 (–17.02). Comparing lines 2 and 3, we
see that rival action speed has a greater negative
effect than rival actions. Line 4 shows the cumula-
tive negative effects of rival action and rival action
speed.

Importantly, even net of the negative effects that
rival actions and rival action speed have on perfor-
mance, the relationship between firm action and
firm performance is still positively sloped. Even
though a Red Queen effect is present in our re-
search, the benefits of focal firm action outweigh
the potentially negative consequences of rival ac-
tion in this competitive contest overall. The same
analysis using ROS to measure performance is also
presented in Figure 3; the results are similar to
those for ROA.

The equations used to generate the plots in Fig-
ure 3 can also be used to calculate the incremental
impact that firm actions, rival actions, and rival
action speed can have on firm performance. In the
case of firm ROA, each additional firm action has
an incrementally positive effect of increasing ROA
by .104 percent while also causing a negative effect
through rival actions and rival action speed that
decreases ROA by .048 percent. Therefore. the net
incremental increase in firm ROA from one firm
action is .056 percent.

Furthermore, using the results from these equa-
tions, it is possible to explore the impact that being
more or less active can have on firm performance.
For example, if we define active firms as those
taking a total number of actions one standard devi-
ation above the mean (23.83 actions) and less active
firms as those taking a total number of actions one
standard deviation below the mean (2.53 actions),
we can make comparisons based on the differential
number of total actions between the two categories
(21.3 actions). The positive effect of those 21.3 ac-
tions is 2.2 or 59.2 percent of the average ROA
(3.74%) in our sample. When we include the neg-
ative effect of those actions that occurs through
rival actions and rival action speed, the results are
still substantial: the net effect on ROA is 1.2 per-
cent, 31.9 percent of the average ROA. Similar re-
sults are found with respect to ROS.

This research contributes by advancing under-
standing of Red Queen competition and the coevo-
lutionary nature of firm search and action, and the
relationship between focal firm action and rival
action on focal firm performance, a key question in
strategy. Specifically, conceiving competition as a
contest of actions, we found support for the Red
Queen effect by theoretically specifying, and em-
pirically detailing, how firm actions and rival ac-
tions have opposing effects on focal firm perfor-

mance. As competitive dynamics is a fairly new
stream of research (Smith et al., 1992), it has lacked
theoretical roots that could give traction to future
research agendas (Smith et al., 2001). The present
study, with its focus on the Red Queen effect, sug-
gests that evolutionary theory may offer important
insights that can advance understanding of the dy-
namics of competition.

In an effort to better understand the boundaries
of the Red Queen hypothesis, we also developed
theory on how Red Queen evolution might depend
upon different industry conditions and market po-
sitions. Importantly, our results indicate that these
factors significantly moderate the effects of firm
actions on rival actions and their joint influence on
performance. We speculated that these moderating
factors block, or facilitate, the learning associated
with search and action for both a focal firm and its
rivals. Our study of these contextual moderating
factors went well beyond antecedent research from
IO economics. That is, despite the central role of
conduct in the structure-conduct-performance par-
adigm, IO researchers testing this theory have fo-
cused mainly on the relationship between structure
and performance and often assumed or not mea-
sured the role of conduct. Additionally, their re-
search has examined the direct effect that industry
context and market position may have on firm per-
formance, to the exclusion of actions or conduct; in
contrast, our study explores the relationship be-
tween actions and performance in the context of
varying concentration, industry demand, and mar-
ket position.

With regard to focal firm performance exhibiting
a more positive relationship to firm performance in
highly concentrated or high-growth industries, our
findings run somewhat counter to our predictions.
We speculate that, in the case of high concentra-
tion, the fact that firm action was only more posi-
tively related to firm performance for one measure
reflects the extent to which firms closely monitor
each other’s actions, are very familiar with each
other’s capabilities and developments, and are so
highly interdependent that they create an environ-
ment in which “surprise” actions are rare, and ri-
vals are more likely to counter actions quickly and
efficiently, wiping out excessive gains. In high-
growth industries, it may be that we didn’t find a
more positive relationship between firm action and
firm performance because firms often act ineffi-
ciently in an effort to keep up with the demands of
the market. Ample demand opportunities may cre-
ate an atmosphere in which firms’ actions are stop-
gaps carried out to meet rising demand without
consideration of their costs. That is, high-demand
environments may create a situation in which firms

2008 75Derfus, Maggitti, Grimm, and Smith

do not have the time to investigate the least costly
way to take action. In this way, firms may waste
resources undertaking actions in haste or perhaps
when they are unnecessary.

With respect to the influence of high concentra-
tion and demand on Red Queen competition, we
found support for the proposition that the relation-
ship between focal firm actions and rival actions is
more intense in highly concentrated industries and
less intense in high-growth industries. These re-
sults support our contention that firms in concen-
trated industries are much more interdependent
than firms in less concentrated industries, while
firms in high-growth industries are less interdepen-
dent than those in low-growth industries. Both re-
sults show how the industry context in which com-
petition takes place moderates the Red Queen effect
on firm evolution and performance.

Market position also appears to have an influ-
ence on Red Queen competition, or the relationship
between focal firm actions, rival actions, and focal
firm performance. Specifically, we predicted that
the positive relationship between a firm’s actions
and its performance would be stronger when the
firm was a market leader, and our findings partially
support this notion. Results for ROS indicate that
firms in stronger leadership positions do receive
greater performance benefits from action. While it
has been suggested that large firms may become
insulated from competitive forces and unrespon-
sive to Red Queen competition (Barnett & McKen-
drick, 2004), our results suggest that large firms
with greater market share can become better com-
petitors and enhance performance by being aggres-
sive with their actions.

Our predictions regarding the influence that mar-
ket leaders’ actions have on rival actions were also
supported. We found that the positive relationships
between focal firm actions and rival actions, and
rival action speed, were weaker when the focal firm
was more of a market leader. We speculate that
either rivals are less likely to act against leading
firms out of fear of retribution, or market leaders
take actions to which it is more difficult for rivals to
respond.

Contrary to our hypothesis, we found that rival
actions have a greater negative impact on the per-
formance of market leaders than on the perfor-
mance of non–market leaders; in essence we found
that “the larger they are, the harder they fall.” This
result may be related to the concept of “judo strat-
egy” as developed by Yoffie and Kwak (2001). With
judo strategy, small rivals can effectively hurt mar-
ket leaders, by eliciting responses that hurt the
market leader more than the rival. When the market
leader’s response affects all customers, it can be

more costly for the leader than the nonleader with
its lower market share. Interestingly, while our
findings did not replicate Barnett and McKen-
drick’s (2004) observation that smaller organiza-
tions were more responsive to Red Queen compe-
tition than larger organizations, we did observe that
smaller firms can be more effective against their
rivals by being aggressive with their actions.

Overall, our findings highlight the intricacies of
the relationship between competition and perfor-
mance and the complexity of studying Red Queen
competition. Though the tests of our baseline hy-
pothesis yielded results that are completely consis-
tent with Red Queen theory, our moderation find-
ings revealed that the effects of search, action, and
learning on firm performance and rival interdepen-
dence largely depend on industry and competitive
context. Further, while context definitely impacts
whether Red Queen competition constrains or en-
hances learning and performance, it does not al-
ways do so in the ways that antecedent research
would suggest. To better understand these relation-
ships, more research needs to be done. As the focus
of this research was on short-term performance,
future research could fruitfully explore longer-term
performance consequences of Red Queen competi-
tion. Are the most active and aggressive firms the
best performers in the long run? How does industry
context influence Red Queen competition over the
long term?

Another potentially fruitful avenue for future re-
search would be to focus on varying action types.
To demonstrate how future research might evolve,
we examined one possible characterization of ac-
tion type, positive sum actions, which may allow
firms to mitigate or reduce the negative aspects of
Red Queen competition. Following Porter’s (1985)
argument that competitors can provide strategic
benefits by helping to develop markets and in-
crease industry demand, some action types, namely
geographic expansions, new marketing campaigns,
and new-product introductions, may represent a
“positive sum competition.” An illustration of this
win-win dynamic can be seen when Starbucks en-
ters a new geographic market in the retail coffee
industry. Rather than negatively affecting the exist-
ing competition in the new market, these competi-
tors often witness increases in their business, as
consumers become more comfortable with spe-
cialty coffee and overall demand increases with the
presence of the new Starbucks (Helliker & Leung,
2002). Similarly, new-product introductions can
positively increase or create demand for all com-
petitors in an industry. For example, Sony’s intro-
duction of the Walkman and Apple’s introduction
of the iPod opened the door for a multitude of

76 FebruaryAcademy of Management Journal

imitations from competitors who garnered reve-
nues from the expanded market.

At least three types of moves are acknowledged
in the literature to have the potential for positive,
demand-expanding effects: geographic expansion,
promotional campaigns, and new-product intro-
ductions.4 To explore the potential of positive sum
actions for future research, we created a measure of
focal firm and rival positive sum actions by sum-
ming actions categorized as geographic, marketing,
and/or product introduction. We calculated firm
positive sum actions by simply summing the num-
ber of instances of all three actions for a focal firm
in a given year. Rival positive sum actions were
operationalized by subtracting a focal firm’s total
number of positive sum actions in a given year from
the total number of positive sum actions taken by
all competitors in the industry.

Post hoc regression of these types of actions
yielded interesting results.5 Specifically, it appears
that focal firm positive sum actions are signifi-
cantly and positively related to firm performance in
the case of return on sales. This result is consistent
with our baseline model. Unlike in our baseline
results, however, here there is a negative and sig-
nificant relationship between firm positive sum ac-
tions and both rival actions and rival action speed.
In addition, rival positive sum actions and firm
performance exhibited no significant relationship.

Taken together, these findings suggest that firm
positive sum actions incite less rivalrous action
and slower rival action speed. Further, when rivals
take positive sum actions, focal firm performance
does not suffer significantly. This finding is again
consistent with the speculation that firms take ac-
tions that build legitimacy and benefit all players in
an industry during the early, high-growth period of
the industry life cycle (Agarwal & Bayus, 2002).
These results also indicate that future research in-
corporating action type could enhance understand-
ing of Red Queen competition. One avenue for fu-
ture research would be to combine action type
effects with an examination of longer-term conse-
quences of Red Queen competition. To clarify, the
above analysis revealed that positive sum actions
may mitigate the negative effects of Red Queen
competition by reducing both the number and
speed of rival actions. However, one could also
conjecture that when a focal firm takes positive
sum actions, rivals are more likely to take positive
sum actions in response, a sequence that may lead
to a rivalry-reducing “loop” with positive perfor-
mance consequences over the longer term. Al-
though a complete analysis is beyond the scope of
our current study, it is interesting to note that there
is a relatively high correlation between firm posi-
tive sum actions and rival positive sum actions
within our data set (r � .41), providing an indica-
tion that firm positive sum actions beget rival pos-
itive sum actions.

This study has implications for practice. Specif-
ically, the study of Red Queen competition pro-
vides a number of insights regarding what actions
managers can take and under what conditions they
should take them in an effort to increase perfor-
mance. Managers of competing firms in highly con-
centrated industries or low-growth industries
should be acutely aware of their mutual interde-
pendence and be cautious of taking actions for fear
of competitive reprisal. Our research also suggests
that challenger firms can effectively hurt industry
leaders by taking action, which is somewhat con-
trary to conventional wisdom. Still, leading firms
received a bigger payoff for acting than did non-
leading firms. Although more research is required,
the post hoc results suggest that managers can po-
tentially avoid the negative consequences of rivalry
by emphasizing positive sum actions such as geo-
graphic actions and product introductions.

Like most research, our study has limitations.
First, although we studied a minimum of 70 per-
cent of the business activity in each of 11 different
industries, our sample favors large, public, single
U.S. business firms that perhaps are in the later
stages of the organizational life cycle. In particular,

4 Geographic expansion, extending a firm’s reach to
customers not previously served, can allow a firm to
avoid head-to-head competition with existing rivals.
Geographic expansion can be targeted where competition
is weak or nonexistent, perhaps in the process filling a
previously underserved geographic segment (Porter,
1980). Promotional campaigns that generate new custom-
ers and new demand are also consistent with a positive
sum notion. As both parties increase their marketing
efforts, the actual or potential customer base can be ex-
panded, creating a situation in which the marketing ac-
tions of the focal firm and the rival firm will both have
positive benefits for the focal firm (Warren, 2002). Mar-
keting campaigns can also aid firms in differentiating
their products. Barney pointed out that product differen-
tiation “reduces the threat of rivalry, because each firm in
an industry attempts to carve out its own unique product
niche” (Barney, 1997: 237). New-product introductions
can also have positive impacts on demand. Each new-
product introduction sets expectations and challenges
rivals to act creatively with their own products in order
to catch up (Schumpeter, 1934). The consequence of
such innovation is that overall demand may increase as
more and better new products are introduced to the mar-
ket over time.

5 Regression results are available upon request from
the authors.

2008 77Derfus, Maggitti, Grimm, and Smith

the competitive dynamics between firms that are
young, small, diversified, and private are not cap-
tured by this study. We speculate that the compet-
itive dynamics among firms in fragmented indus-
tries and in early stages of the life cycle, where
mutual interdependence is lower, would be quite
different. Future research should continue investi-
gation of these issues in these other industry con-
texts. As is the case in competitive dynamics re-
search, our study also captures only observable
moves reported in the press and publications we
examined. In addition, our use of product-market-
based industries to define the competitive land-
scape may not address the potentially changing
nature of competition. For example, it is likely that
focal firms in our sample face competitors from
outside their industry or outside the United States
that offer substitute products, have similar resource
positions, or exist in geographic markets that are
new to the focal firms (Chen, 1996; Peteraf & Ber-
gen, 2003). Future research can overcome this lim-
itation by identifying competitors on the basis of
criteria other than product-market commonality.

Additionally, although we theorize that search
and learning take place in this Red Queen com-
petitive context, we do not directly specify and
measure these constructs. Future research could
further elucidate Red Queen competition by ex-
amining more specifically the search and learn-
ing that take place. Future research could also
consider additional links beyond the scope of the
current study, such as whether and how firm
competitive experience and performance shape
future firm action. We can speculate that a firm’s
prior learning and experience with action affect
its future action and action speed. Moreover, ad-
ditional research could examine the extent to
which a focal firm learns vicariously from other
firms (Baum, Li, & Usher, 2000).

In conclusion, the most important contribution
of this research is its theoretical and empirical
examination of Red Queen competition, con-
ceived of as the positive and negative conse-
quences of firm actions on performance. Specifi-
cally, the research shows that firm actions can
play out as a Red Queen race among rivals: Firm
actions are related to rival actions and rival ac-
tion speed, and all can impact focal firm perfor-
mance. However, we also demonstrated that the
effects of firm action on performance are complex
and dependent on the industry context and the
market positions of competitors. Additional the-
ory and research are needed to improve under-
standing of Red Queen competition.

REFERENCES

Agarwal, R., & Bayus, B. 2002. The market evolution and
take-off of new product innovations. Management
Science, 48:1024 –1041.

Agarwal, R., & Gort, M. 1996. The evolution of markets
and entry, exit and survival of firms. Review of
Economics and Statistics, 78: 489 – 498.

Agarwal, R., Sarkar, M. B., & Echambadi, R. 2002. The
conditional effect of time on firm survival. Academy
of Management Journal, 45: 971–994.

Argote, L. 1999. Organizational learning: Creating, re-
taining and transferring knowledge. Boston: Klu-
wer.

Bain, J. S. 1951. Relation of profit rate to industry con-
centration: American manufacturing, 1936 –1940.
Quarterly Journal of Economics, 65: 293–324.

Barnett, W. P., & Hansen, M. T. 1996. The Red Queen in
organizational evolution. Strategic Management
Journal, 17: 139 –157.

Barnett, W. P., & McKendrick, D. 2004. Why are some
organizations more competitive than others? Evi-
dence from a changing global market. Administra-
tive Science Quarterly, 49: 535–571.

Barnett, W. P., & Sorenson, O. 2002. The Red Queen in
organizational creation and development. Industrial
and Corporate Change, 11: 289 –325.

Barney, J. 1986. The types of competition and the theory
of strategy: Toward an integrative framework. Acad-
emy of Management Review, 11: 791– 800.

Barney, J. 1997. Gaining and sustaining competitive
advantage. Reading, MA: Addison-Wesley.

Baum, J. A. C., Li, S. X., & Usher, J. M. 2000. Making the
next move: How experiential and vicarious learning
shape the locations of chains’ acquisitions. Admin-
istrative Science Quarterly, 45: 766 – 801.

Baumol, W. J. 2004. Red-queen games: Arms races, rule
of law and market economies. Journal of Evolution-
ary Economics, 14: 237–247.

Bothwell, J. T., Cooley, T., & Hall, T. 1984. A new view of
the market structure-performance debate. Journal of
Industrial Economics, 32: 397– 417.

Bradburd, R. M., & Caves, R. 1982. A closer look at the
effect of market growth on industries profits. Re-
view of Economics and Statistics, 64: 635– 645.

Cameron, A., & Travendi, P. 1986. Economic models
based on count data: Comparisons and applications
of some estimators and tests. Journal of Applied
Econometrics, 1: 29 –53.

Carroll, G., & Hannan, M. 1989. Density dependence in
the evolution of newspaper organizations. American
Sociological Review, 54: 524 –541.

Carroll, G., & Hannan, M. 2000. The demography of
corporations and industries. Princeton, NJ: Prince-
ton University Press.

78 FebruaryAcademy of Management Journal

Carroll, L. 1960. The annotated Alice: Alice’s adven-
tures in wonderland and through the looking-
glass. New York: New American Library.

Caves, R. 1980. American industry: Structure, conduct,
performance (6th ed.). Englewood Cliffs, NJ: Pren-
tice-Hall.

Caves, R., & Porter, M. E. 1977. From entry barriers to
mobility barriers. Quarterly Journal of Economics,
91: 241–261.

Chen, M. J. 1988. Competitive strategic interaction: A
study of competitive actions and responses. Un-
published manuscript, University of Maryland, Col-
lege Park.

Chen, M. J. 1996. Competitor analysis and interfirm ri-
valry: Toward a theoretical integration. Academy of
Management Review, 21: 100 –135.

Chen, M. J., & Miller, D. 1994. Competitive attack, retal-
iation, and performance: An expectancy-valence
framework. Strategic Management Journal, 15: 85–
102.

Comanor, W., & Wilson, T. 1967. Advertising, market
structure, and performance. Review of Economics
and Statistics, 49: 423– 440.

Cubbin, J., & Domberger, S. 1988. Advertising and post-
entry oligopoly behaviour. Journal of Industrial
Economics, 37: 123–140.

Dawkins, R. & Krebs, J. R. 1979. Arms races between and
within species. Proceedings of the Royal Society of
London, B205: 489 –511.

Eisenhardt, K. M., & Tabrizi, B. N. 1995. Accelerating
adaptive processes: Product innovation in the global
computer industry. Administrative Science Quar-
terly, 40: 84 –110.

Erez, A., Bloom, M. C., & Wells, M. T. 1996. Using ran-
dom rather than fixed effects models in meta-analy-
sis: Implications for situational specificity and valid-
ity generalization. Personnel Psychology, 49: 275.

Ferrier, W. J., Smith, K. G., & Grimm, C. M. 1999. The role
of competitive action in market share erosion and
industry dethronement: A study of industry leaders
and challengers. Academy of Management Journal,
42: 372–383.

Frome, E. L., Kutner, M. H., & Beauchamp, J. J. 1973.
Regression analysis of Poisson-distributed data.
Journal of American Statistical Association, 68:
935–940.

Gaskins, D. W. 1971. Dynamic limit pricing: Optimal
limit pricing under threat of entry. Journal of Eco-
nomic Theory, 3: 306 –322.

Gilbert, R. J., & Newberry, D. M. 1982. Preemptive pat-
enting and the persistence of monopoly. American
Economic Review, 72: 514 –526.

Graham, E. M. 1979. Technological innovation and the
dynamics of the U.S. comparative advantage in in-
ternational trade. In C. T. Hill & J. M. Utterback

(Eds.), Technological innovation for a dynamic
economy. New York: Pergamon.

Greene, W. H. 1993. Econometric analysis (2nd ed.).
New York: MacMillan.

Grimm, C. M., & Smith, K. G. 1997. Strategy as action:
Industry rivalry and coordination. Cincinnati:
South-Western College Publishing.

Helliker, K., & Leung, S. 2002. Counting beans: Despite
the jitters, most coffeehouses survive Starbucks.
Wall Street Journal, September 24, 2002: A1.

Ingram, P. 2002. Interorganizational learning. In J. A. C.
Baum (Ed.), Companion to organizations: 642– 664.
Oxford, UK: Blackwell.

Jauch, L. R., Osborn, R. N., & Martin, T. N. 1980. Struc-
tured content analysis of cases: A complimentary
method for organizational research. Academy of
Management Journal, 5: 517–526.

Kamien, M., & Schwartz, N. 1971. Limit pricing and
uncertain entry. Econometrica, 39: 441– 454.

Levitt, B., & March, J. G. 1988. Organizational learning. In
W. R. Scott (Ed.), Annual review of sociology, vol.
14: 319 –340. Palo Alto, CA: Annual Reviews.

Liebowitz, S. J. 1982. Measuring industrial disequilibria.
Southern Economic Journal, 49: 119 –136.

March, J., & Simon, H. 1958. Organizations (2nd ed.).
Cambridge, MA: Blackwell.

Masson, R. T., & Shannan, J. 1986. Excess capacity and
limit pricing: An empirical test. Economica, 53:
365–378.

Miller, D., & Chen, M. J. 1994. Sources and consequence
of competitive inertia: A study of the U.S. airline
industry. Administrative Science Quarterly, 39:
1–23.

Nelson, R. R., & Winter, S. G. 1982. An evolutionary
theory of economic change. Cambridge, MA:
Belknap.

Peteraf, M. A., & Bergen, M. E. 2003. Scanning dynamic
competitive landscapes: A market-based and re-
source-based framework. Strategic Management
Journal, 24: 1027–1041.

Pisano, G. P. 1994. Knowledge, integration and the locus
of learning: An empirical analysis of process devel-
opment. Strategic Management Journal, 15: 85–
100.

Porter, M. E. 1980. Competitive strategy. New York: Free
Press.

Porter, M. E. 1985. Competitive advantage: Creating
and sustaining superior performance. New York:
Free Press.

Scherer, F. M., & Ross, D. 1990. Industrial market struc-
ture and economic performance. Boston: Hough-
ton-Miller.

Schmalensee, R. 1976. Advertising and profitability: Fur-

2008 79Derfus, Maggitti, Grimm, and Smith

ther implications of the null hypothesis. Journal of
Industrial Economics, 25: 45–54.

Schomburg, A. J., Grimm, C. M., & Smith, K. G. 1994.
Avoiding new product warfare: The role of industry
structure. In P. Shrivastava, A. Huff, & J. Dutton
(Eds.), Advances in strategic management, vol. 10:
145–174. Greenwich, CT: JAI.

Schumpeter, J. 1934. The theory of economic develop-
ment. Cambridge, MA: Harvard University Press.

Schumpeter, J. A. 1976. Capitalism, socialism, and de-
mocracy (5th ed.). London: George Allen & Unwin.

Smith, K. G., Ferrier, W. J., & Ndofor, H. 2001. Competi-
tive dynamics research: Critique and future direc-
tions. In M. A. Hitt, R. E. Freeman, & J. S. Harrison
(Eds.), Handbook of strategic management: 315–
361. Oxford, UK: Blackwell.

Smith, K. G., Grimm, C. M., & Gannon, M. J. 1992. Dy-
namics of competitive strategy. London: Sage.

Spence, A. M. 1977. Entry, capacity, investment and
oligopolistic pricing. Bell Journal of Economics, 8:
531–544.

van Valen, L. 1973. A new evolutionary law. Evolution-
ary Theory, 1: 1–30.

Waldman, D. E., & Jensen, E. J. 2001. Industrial organi-
zation. Boston: Addison-Wesley.

Warren, K. 2002. Competitive strategy dynamics. New
York: Wiley.

Weick, K. 1995. Sensemaking in organizations. Thou-
sand Oaks, CA: Sage.

Williamson, O. E. 1965. A dynamic theory of interfirm
behavior. Quarterly Journal of Economics, 79: 579 –
607.

Yoffie, D. B., & Kwak, M. 2001. Judo strategy. Boston:
Harvard Business School Publishing.

Young, G. 1993. Engaging a rival: Industry and firm
specific predictors of rivalrous firm conduct in the
U.S. software industry. Unpublished manuscript,
University of Maryland, College Park.

Young, G., Smith, K. G., & Grimm, C. M. 1996. “Austrian”
and industrial organization perspectives on firm-
level competitive activity and performance. Organi-
zation Science, 7: 243–254.

Pamela J. Derfus (pderfus@verizon.net) received her
Ph.D. in strategic management from the Robert H. Smith
School of Business at the University of Maryland. Her
research interests lie in the areas of competitive dynam-
ics, strategy implementation, and cooperation between
firms.

Patrick G. Maggitti (maggitti@temple.edu) is an assistant
professor of management and entrepreneurship in the
Fox School of Business at Temple University. His re-
search focuses on the dynamics of competition and the
decision making of executives, entrepreneurs, and inves-
tors. He received his Ph.D. in strategic management from
the Robert H. Smith School of Business at the University
of Maryland.

Curtis M. Grimm (cgrimm@rhsmith.umd.edu) is the
Dean’s Professor of Supply Chain and Strategy at the
Robert H. Smith School of Business, University of Mary-
land. He received his Ph.D. in economics from the Uni-
versity of California, Berkeley, with primary focus on
industrial organization. Professor Grimm’s research has
focused on the interface of business and public policy
with strategic management, with a particular emphasis
on competition and competition policy.

Ken G. Smith (kgsmith@rhsmith.umd.edu) is the Dean’s
Chair and a professor of business strategy at the Robert H.
Smith School of Business, University of Maryland. He
earned a Ph.D. in business policy from the University of
Washington. His research interests include strategic de-
cision making, competitive dynamics, and the manage-
ment of knowledge and knowledge creation.

80 FebruaryAcademy of Management Journal

r Academy of Management Journal
2017, Vol. 60, No. 5, 1882–1914.
https://doi.org/10.5465/amj.2015.029

5

RED QUEEN COMPETITIVE IMITATION IN THE U.K. MOBILE
PHONE INDUSTRY

CLAUDIO GIACHETTI
Ca’ Foscari University of Veni

ce

JOSEPH LAMPEL
University of Manchester

STEFANO LI PIRA
University of Warwick

This paper uses Red Queen competition theory to examine competitive imitation. We
conceptualize imitative actions by a focal firm and its rivals along two dimensions:
imitation scope, which describes the extent to which a firm imitates a wide range (a

s

opposed to a narrow range) of new product technologies introduced by rivals; and
imitation speed, namely the pace at which it imitates these technologies. We argue that
focal firm imitation scope and imitation speed drive performance, as well as imitation
scope and speed decisions by rivals, which in turn influence focal firm performance. We
also argue that the impact of this self-reinforcing Red Queen process on firms’ actions
and performance is contingent on levels of product technology heterogeneity—defined
as the extent to which the industry has multiple designs, resulting in product variety. We
test our hypotheses using imitative actions by mobile phone vendors and their sales
performance in the U.K. from 1997 to 2008.

Once we become self-consciously aware that the
possibilities of innovation within any one company
are in some important ways limited, we quickly see
that each organization is compelled by competition to
look to imitation as one of its survival and growth
strategies. (Levitt, 1966: 38)

The emergence of what has often been referred to
as the “new economy” has greatly expanded re-
search on the power of technological innovation to
create competitive dynamics that can reshape in-
dustries (Baumol, 2004; Teece, 1998). While the fo-
cus on innovation as the engine of industry evolution
reflects both the potential gains that accrue to first
movers (Lieberman & Montgomery, 1988), and the

dramatic impact of disruptive technologies on the
competitive landscape (Christensen & Bower, 1996),
it inadvertently tends to eclipse the importance of
imitation as an agent of change. Researchers that take
a broader perspective see imitation as the twin pro-
cess to innovation that, arguably like innovation,
also plays a role in industry evolution in all contexts
(Cohen & Levinthal, 1989; Levitt, 1966; Semadeni &
Anderson, 2010), but takes on even greater signifi-
cance in the rapidly changing technology-intensive
industries that constitute the new economy. As
Baumol (2004: 246–247) observed,

in the new economy no firm [. . .] can afford to fall
behind its rivals. […] If a firm fails to adopt the latest
technology—even if the technology is created by
others—then its rivals can easily take the lead and
make disastrous inroads into the slower firm’s sales.

Formulating an effective imitation strategy is
a problem that confronts managers in any industry
(Lieberman & Asaba, 2006), but in industries with
rapid technological change the problem is com-
pounded by higher levels of uncertainty about the
market performance of new product technologies
(Utterback & Suarez, 1993). This “technological

We would like to thank Associate Editor Dovev Lavie
and three anonymous reviewers for their invaluable com-
ments and guidance during the review process, which
helped strengthen this article. We would also like to thank
Marco Li Calzi, Massimo Warglien, Francesco Zirpoli,
Anna Comacchio, Juan Pablo Maicas, and Gianluca Marchi
for their insightful comments on earlier drafts of this arti-
cle. Finally, we thank seminar participants at Ca’ Foscari
University of Venice, and at a 2015 Academy of Manage-
ment Conference paper session, for their astute remarks on
earlier versions of this article.

188

2

Copyright of the Academy of Management, all rights reserved. Contents may not be copied, emailed, posted to a listserv, or otherwise transmitted without the copyright holder’s express
written permission. Users may print, download, or email articles for individual use only.

https://doi.org/10.5465/amj.2015.0295

uncertainty” presents managers with considerable
challenges when deciding how far and how fast they
should imitate their rivals, and this challenge per-
sists when their decisions, in turn, create competi-
tive conditions that may bring further pressure to
imitate (Gaba & Terlaak, 2013; Rhee, Kim, & Han,
2006). In this paper, we address the questions of the
extent to which, and speed with which, firms should
imitate their rivals, taking into account both the com-
petitive dynamics that ensue as a result of innovation
and imitation decisions, and the level of technological
uncertainty in rapidly changing technology-intensive
industries.

Our analysis of imitation must begin with the
recognition that imitation is a strategic choice that
firms pursue when they wish to lower risks and
costs by learning from their rivals’ actions, espe-
cially when these actions involve pioneering new
technologies, or launching radically innovative
products (Ethiraj & Zhu, 2008; Lee, Smith, Grimm, &
Schomburg, 2000). However, imitation is also
a competitive move that can pose a threat to rivals
that have yet to adopt the pioneering technologies,
or introduce products with similar features. When
some of the laggards react to the threat by also imi-
tating, this gives rise to “competitive imitation,”
a process in which imitation by some of the firms in
an industry puts competitive pressure on the rest to
also imitate. This process is consistent with the re-
lationship between action and reaction that has
been extensively studied by competitive dynamics
research (Smith, Ferrier, & Ndofor, 2001). The main
premise of competitive dynamics is that the actions
of one firm, or group of firms, trigger reactions by
other firms, which in turn produce a series of ac-
tions and reactions that continue as long as firms
seek to improve their competitive position (Derfus,
Maggitti, Grimm, & Smith, 2008; Smith, Grimm,
Gannon, & Chen, 1991). In technology-intensive
industries, innovation triggers the competitive im-
itation process. Faced with mounting evidence that
the innovator’s new product technologies are find-
ing a market, other firms that previously refused to
match the innovator’s move begin to experience
increasing pressure to imitate. Their imitative move
serves to entrench the new technologies in the
market even further, which in turn not only in-
creases competitive imitation but also accelerates
the evolution of the industry.

The coevolutionary process by which firms act
and react to each other has been shown by organi-
zational scholars to influence both firm performance
and industry structure (Nelson & Winter, 1982). Not

unexpectedly, scholars have also noted that co-
evolutionary processes in competitive environments
have parallels with biological evolution. The paral-
lels have led scholars to borrow from the work of
evolutionary biologists, notably Van Valen’s (1973)
work on the coevolution of dynamically interacting
species. Of particular interest is the “Red Queen”
effect, the allusion made by Van Valen (1973) to
Alice’s encounter with the Red Queen in Lewis
Carroll’s Through the Looking-Glass (Carroll, 1960),
when he sought to explain the constant probability of
species’ extinction regardless of the duration of their
evolutionary history.1 Organization researchers have
argued that what holds for biological evolution is
in principle also the case in business contexts.
Thus, firms can be said to engage in a “Red Queen
competition” (Barnett & Hansen, 1996; Barnett &
Sorenson, 2002); that is, the continuous and esca-
lating activity of firms trying to maintain relative
fitness in a dynamic system, such that they end up
improving as fast as they can just to stand still rela-
tive to competitors.

We draw on the literature on Red Queen compe-
tition, and research on competitive dynamics, imi-
tation, and technology innovation, to build a model
that captures how decisions to imitate new product
technologies stimulate further imitation by rivals,
and how this “competitive imitation” in turn in-
fluences, and is influenced by, changing industry
conditions. Our study complements the competitive
dynamics and imitation literature in several re-
spects. To begin with, most extant research on imi-
tation of innovations has tended to see imitation as
a binary variable: firms either imitate or they do
not (e.g., Greve, 1998; Hsieh & Vermeulen, 2013;
Makadok, 1998). In practice firms seldom imitate, or
do not imitate, every aspect of their rivals’ offerings,
but instead tend to imitate some of the features of
products introduced by rivals, while retaining
existing features (Bayus & Agarwal, 2007; Giachetti &
Dagnino, 2017). From this it follows that managers
face two basic questions when they consider imita-
tion as the best next move: the first is how much to
copy, i.e., “imitation scope” (Csaszar & Siggelkow,
2010; Narasimhan & Turut, 2013), and the second is

1 In reference to Carroll’s tale, when the Red Queen re-
sponds to Alice “here, you see, it takes all the running you
can do, to keep in the same place” (Carroll, 1960: 345), Van
Valen noted that biological evolution features such
change: species must constantly adapt in order to survive,
while confronting ever-evolving rival species in an ever-
changing environment.

2017 1883Giachetti, Lampel, and Li Pira

how quickly to imitate, i.e., “imitation speed” (Lee
et al., 2000). In this paper, we argue that the scope
and speed decisions of one firm influence the scope
and speed decisions of rivals. Rivals’ imitation
scope and speed decisions will then influence the
firm’s subsequent scope and speed decisions. What
applies to the reaction of rivals to the actions of
a single firm is also true for the industry as a whole. If
we step one level of analysis up to consider the entire
group of industry rivals, we can see Red Queen
competition from a wider perspective: the speed and
scope decisions made by firms at different times in-
duce each other’s speed and scope decisions. Our
contribution in this paper is to show that Red Queen
competition in technology-intensive industries es-
calates the magnitude of imitation speed and scope
choices for all competitors.

Second, thus far, competitive dynamics studies
have not examined how changes in the techno-
logical environment may affect the Red Queen
cycle. This analysis is especially important in
technology-intensive industries where technolog-
ical change, for example the emergence or decline
of dominant designs, can dramatically alter com-
petition (Chen & Turut, 2013), render obsolete
a firm’s capabilities (Barkema, Baum, & Mannix,
2002; Bayus & Agarwal, 2007; Utterback & Suarez,
1993), and encourage firms to develop a new
technology imitation strategy (Narasimhan &
Turut, 2013). It is not difficult to see that changes
in the technological environment that drive new
product introductions are often central to the types
of moves that drive Red Queen competition. For
example, recent studies in the imitation and tech-
nology innovation literature (Argyres, Bigelow, &
Nickerson, 2015; Giachetti & Lanzolla, 2016;
Madhok, Li, & Priem, 2010; Posen, Lee, & Yi, 2013)
have shown that as industries evolve, changes in
technologies, and subsequently their diffusion,
can influence rates of imitation. These studies
complement prior work by authors such as
Utterback and Suarez (1993), who pointed out that
as the industries mature they tend to transition
from high to low levels of product technology
heterogeneity—where low levels of product tech-
nology heterogeneity correspond to the emergence
of design dominance. Low product technology
heterogeneity in turn leads to low technological
uncertainty: firms find it easier to know which
design options are more likely to yield good market
performance, and are thus more likely to imitate.
Our paper examines how product technology het-
erogeneity moderates the Red Queen effect.

Third, existing competitive dynamics studies
have tended to examine antecedents and perfor-
mance outcomes of action types such as pricing,
marketing, and capacity expansion, in industries
such as professional services, professional sports,
and motion pictures, where technology is a minor
competitive factor (Lampel & Shamsie, 2009; Ross &
Sharapov, 2015; Semadeni & Anderson, 2010); or in
industries such as airlines, where technology is
more important but is still peripheral to the main
factors responsible for success (Chen & Miller, 1994;
Miller & Chen, 1994; Smith et al., 1991). In contrast,
we chose to examine Red Queen competitive dy-
namics in an industry where “creative destruction”
(Schumpeter, 1942), triggered by the introduction
of new products and technologies, is the primary
competitive force. The mobile phone industry is
a rapidly changing technology-intensive industry
where continuous and swift imitation of rivals’ in-
novation is a key prerequisite for handset vendors to
maintain competitive parity. More specifically, our
research site is the U.K. mobile phone industry from
1997 to 2008, a period during which the industry
evolved rapidly, driven by incessant rivalry among
a dozen handset vendors to get or keep ahead of one
another.

The rest of our paper is structured as follows. We
begin with an overview of Red Queen theory. We
subsequently define and discuss imitation scope,
imitation speed, and product technology heteroge-
neity, and derive hypotheses about how these as-
pects influence Red Queen competitive imitation.
We then describe our methods and present our re-
sults. We conclude with limitations of our study and
suggestions for future research.

THEORY BACKGROUND AND HYPOTHESES

Red Queen Competitive Imitation: Focal Firm,
Rivals, and Firm Performance

In this section, we develop a theory that explains
the Red Queen effect in terms of a firm’s imitation of
new product technologies, rivals’ imitation of new
product technologies, and their combined impact on
the firm’s performance. To ensure that our theory
development is consistent and clear, it is important
to define imitation in contrast to innovation before
we move forward. As pointed out by Semadeni and
Anderson (2010), in markets where firms can closely
examine their competitors’ product offerings and
track the market performance of those offerings in
real time, firms can choose between introducing to

1884 OctoberAcademy of Management Journal

the market products with new features, or confining
their actions to copying features previously in-
troduced by rivals. We likewise also distinguish be-
tween introduction and copying, and define
innovation as introducing first to the market prod-
ucts that contain new features, and imitation as
copying others’ innovations.

In this paper we focus our analysis on imitative
actionswhile controlling for innovative actions. Two
types of imitative decisions are examined: imitation
scope and imitation speed. Our decision is based on
evidence provided by different but complementary
streams of literature. On the one hand, the technol-
ogy innovation literature has argued that a wider
imitation scope is an indication that the firm’s
products can stay abreast of new technologies
(Narasimhan & Turut, 2013). Yet, at the same time,
the literature on first-mover advantage, as well as
competitive dynamics literature, has focused more
closely on imitation speed, arguing that higher imi-
tation speed is a signal the firm is one of the first
players committed to adopting new technologies so
as to keep up with innovators and differentiate with
respect to laggard rivals (Lee et al., 2000; Markides &
Geroski, 2004). Though these studies have been in-
terested in whether imitation represents a source of
performance differences, they have posed somewhat
different research questions, and thus have pro-
gressed along independent trajectories. In practice,
when a firm faces a group of rivals who are in-
troducing new products with a variety of features at
different times, they cannot focus only on scope or
speed, but must consider both the question of how
many of the features the firm should imitate, and also
how quickly it should proceed with imitation. In this
article, we propose to bring together the different
analyses of imitation and competition explored in
these bodies of literature in order to obtain a broader
understanding of the roles that imitation scope and
speed play in sustaining the Red Queen competitive
imitation.

Because Red Queen competition describes a re-
ciprocal back-and-forth process, firms play different
roles in different time periods, and it is important to
be clear and consistent about the labels we use when
referring to firms. In Red Queen papers, the “focal
firm” and “rivals” may switch places in the analysis
over time. The “focal firm” is the industry player
whose imitative moves attract attention and call for
a response from other firms, the “rivals” at that spe-
cific point in time. For example, we could say that at
time t the “focal firm,” having observed new product
technologies introduced by one or several of its

“rivals” in t 2 1, must decide how many of these
technologies it should imitate. In this context, “ri-
vals” are all the other firms within the industry that
the “focal firm” sees as competitors. “Rivals,” for
their part, observe the focal firm’s moves, gauge the
resulting performance, and decide on how many of
these moves they should imitate at time t 1 1. This
turns the rival firms into focal firms, who are now
observing and analyzing moves recently made by
rivals. Their actions challenge rivals, who must now
consider their moves, and so on.

To summarize, the baseline Red Queen competi-
tive imitation we develop in this section works as
follows. Focal firms that successfully imitate new
product technologies obtain performance advantages
(e.g., sales increases) by virtue of competitive advan-
tage that they hold vis-à-vis rivals that imitate either
less intensively (i.e., lower imitation scope), or more
slowly (i.e., lower imitation speed). Higher perfor-
mance of focal firms that imitate more intensively, or
more rapidly, combined with performance losses ex-
perienced byrivals, willmotivatethe lattertorespond
by increasing their imitation scope and speed. The
more intense and rapid the rivals’ imitative response,
the more the focal firm experiences a threat to its
performance, and the more it feels pressure to
respond—by innovating or imitating.

It is worth noting that our theory of Red Queen
competitive imitation describes competition as the re-
sult of a sequence of imitative actions after a set of
new product technologies are introduced. We argue
that focal firms imitate innovators (i.e., technology pi-
oneers), and rivals subsequently imitate focal firms
in an incessant race to maintain competitive parity
(Lieberman&Asaba,2006).Morespecifically,whilethe
rationale for the first imitations (by the quickest imita-
tors) is “informationally based”—i.e., when making
imitative decisions first imitators use the information
generated by market performance of the new technol-
ogies introduced by innovators—the rationale for
subsequent imitations is also motivated by “competi-
tive bandwagon” pressure (Abrahamson & Rosenkopf,
1993)—i.e., the pressure on nonimitators when they
face diminishing profit opportunities as more of their
competitors imitate innovative first movers.

The Competitive Advantage of More Active Firms:
Learning and Repertoires of Actions

Taking their inspiration from Joseph Schumpeter,
specifically his concept of “creative destruction”
(Schumpeter, 1942)—which, concisely summarized,
arguesthat competition is adynamicmarket processin

2017 1885Giachetti, Lampel, and Li Pira

which entrepreneurs trigger and respond to change—
competitive dynamics research has shown that more
“active” firms, defined as those that take more frequent
competitive actions than most of their industry rivals,
are more likely to attain higher performance (Ferrier,
Smith, & Grimm, 1999; Young, Smith, & Grimm, 1996).
In contrast, firms that lag behind most of their industry
rivalswhen it comestotaking competitiveactionstend
to be at a competitive disadvantage (Miller & Chen,
1994). There are several related factors that account for
this relationship. First, firms that are more active are
more likely to keep pace with change in a rapidly
changing environment (Chen, Lin, & Michel, 2010;
Ndofor, Sirmon, & He, 2011; Smith, Ferrier, & Ndofor,
2001). Second, because they make more moves, these
firms are also more likely to take actions that change
the environment in ways that are favorable to them,
and lessfavorabletolessactivefirms (Rindova, Ferrier,
& Wiltbank, 2010). Finally, in dynamic environments
in which the direction and consequences of change are
uncertain, firms that are more active have a shorter
learning cycle compared to firms that are less active.
Active firms capture and put to use the knowledge
gained from observing their rivals more quickly com-
pared to firms that hesitate (Baum & Ingram, 1998;
Baum, Li, & Usher, 2000; Greve, 1996).

Learning also plays a central role in research on
Red Queen competition. Initial Red Queen studies
sought to show that competition and learning trigger
one another in an ongoing, self- reinforcing process
(Barnett & McKendrick, 2004; Barnett & Sorenson,
2002). As Barnett and Sorenson (2002: 290) put it,
Red Queen is a process that results when “competi-
tion among organizations triggers internal learning
processes; and learning increases the strength of
competition generated by an organization.” More
recent Red Queen research has focused to a greater
extent on learning as a process in which rivals try to
figure out the causal mechanism that links a reper-
toire of competitive moves to performance (Derfus
et al., 2008). The simplest competitive repertoire
consists of a single move. In markets where a single
move type is central to performance (e.g., price re-
duction), Red Queen is confined to single-type tit-
for-tat responses. In most markets, however, the focal
firm’s competitive advantage (or disadvantage) re-
sults from a combination of successful (or failed)
competitive actions,2 and firms face choices about
which combination of moves they should employ. If

the repertoire of possible moves focuses primarily on
product technologies, firms have to assess which of
the new technologies launched by rivals should be
imitated, and which should be avoided.

In the remaining part of this theory section, we
develop a set of hypotheses about our theory of Red
Queen competitive imitation. Our argument is that
focal firms will perform better than “less active”
imitators if they are “more active,” both in terms of
the number of new product technologies they imitate
and the speed at which they are able to imitate.
Further, we argue that product technology heteroge-
neity may constrain focal firms’ learning capabilities,
obstructing their ability to increase performance via
imitative actions.

Scope and Average Speed of a Firm’s Imitation of
New Product Technologies and its Performance

How much to copy: Imitation scope as a com-
petitive response. In the specific context of new
product technology in which we are interested,
multiple imitation opportunities present firms with
the strategic choice regarding how many of the
technologies introduced by rivals they should
imitate. This scenario is typical in technology-
intensive industries, such as consumer electronics
(e.g., mobile phones and personal computers),
where firms constantly face competitive threats
from new product technologies that expand the set
of functionalities that are offered to consumers
(Bayus & Agarwal, 2007). The choice that confronts
firms as new products with new functionalities
enter the market is how many of these functional-
ities they should incorporate into their products.
The choice targets what we call “imitation scope;”
that is, the extent to which a firm (in a given period)
imitates a wide number (as opposed to a narrow
number) of new product technologies introduced
by competitors.

When looking at imitation scope, we have to bear
in mind that consumers evaluate the desirability of
adopting new features in the context of the entire
bundle of functionalities offered by the product
(O’Shaughnessy, 1989). In other words, consumers
compare products with, and without, a given func-
tionality before making a purchase. The inclusion of
a functionality will not necessarily motivate them to
make a purchase, unless the additional functionality
adds to the value of the package as a whole. First
movers (i.e., innovators) must make this evaluation
without prior market data (or at best consumer re-
search data), while imitators can use the market

2 See Chen and Miller (2012) for an extensive review
comparing studies on single actions versus action
repertoires.

1886 OctoberAcademy of Management Journal

performance of new functionalities when making
this decision (Carpenter & Nakamoto, 1989). The
problem, however, is that firms have data on mul-
tiple functionalities. Some of these functionalities
are present in the same product, which makes it
difficult to evaluate them separately, while other
functionalities are spread across multiple products
and present in a variety of combinations—creating
an even greater evaluation challenge (Krishnan &
Bhattacharya, 2002).

If firms cannot analyze the sales potential of indi-
vidual functionalities, the question that arises is
whether they can evaluate the potential of sets of
functionalities. Technology innovation literature that
has examined the consumer buying behavior of
products with multiple functionalities (Chen & Turut,
2013) has suggested that when firms have to assess
how consumers evaluate a set of objects—in our case,
products that offer certain functionalities—they will
evaluate the options they are presented by consider-
ing both the absolute utility of each feature (e.g., text
messaging in mobile phones), and their relative
standing in the choice set (i.e., how valuable text
messaging is relative to other functionalities in the
set). The evaluation relies on reference points that are
endogenous to the choice set (Baucells, Weber, &
Welfens, 2011). This can be the product that the
consumer currently owns, or some idealized combi-
nation of functionalities in the product that the con-
sumer wishes to purchase (Zhou, 2011). Reference
points in a technologically mature industry where
products perform a stable set of well-established
functionalities are more likely to be based on price,
since the difference between the functionalities of old
and new products is not substantial. However, in in-
dustries where technology is evolving rapidly, as in
most technology-intensive industries, consumers’
reference points are future oriented, and tend to
change as new functionalities are introduced. As
Chen and Turut (2013: 2748) put it:

Context dependent preferences are especially rele-
vant for consumers’ adoption of technology in-
novation because the reference points of product
attributes in consumers’ minds are likely to evolve
over time with the advance of technology and the ar-
rival of new products in the market; this influences
consumers’ adoption of products with new technol-
ogy and consequently firms’ innovation strategies.

Introduction of new functionalities in the form of
new product features or attributes tends to shift the
reference point toward the innovative feature, and
away from old features. Put differently, consumers

will value the entire set of functionalities in a prod-
uct more if the product includes new functionalities
that represent the next step in the evolution of un-
derlyingtechnologies. This shift in reference point as
technology evolves strongly influences the compet-
itive logic in these markets. While it creates in-
centives to innovate new functionalities, it creates
even stronger incentives to imitate (Narasimhan &
Turut, 2013).

Narasimhan and Turut (2013) provided empirical
support for the advantages of imitation, showing that
firms attain higher performance if they choose to im-
itate as many pioneering features introduced by rivals
as possible, rather than differentiate by introducing
their own features. Their conclusions are in line with
other empirical studies of consumer attitudes sug-
gesting that in markets where technology is rapidly
evolving, consumers evaluate more favorably brands
with a reputation for staying abreast of new technol-
ogies, while at the same time displaying a strong bias
against brands that lack the latest technologies
(O’Shaughnessy, 1989; Pessemier, 1978). From the
point of view of firms that are considering how many
of the new functionalities they should adopt in the
new product offerings, this suggests that firms are
more likely to gain sales if they adopt as many of the
new features as their capabilities will allow. This
leads to the following hypothesis:

Hypothesis 1a. An increase in the focal firm
scope of imitation of new product technologies
will positively influence its performance.

How fast to copy: Average speed of imitation as
a competitive response. Another question firms
must confront is how quickly to imitate rivals’ moves
(Markides & Geroski, 2004). Similar to our discus-
sion on imitation scope in technology-intensive in-
dustries where firms launch products that combine
multiple technologies, and hence present multiple
imitation opportunities, a related decision that con-
fronts firms is how quickly these multiple technol-
ogies should be imitated. At the product line level,
this choice targets what we call “average speed of
imitation:” the average time it takes for the focal firm
to adopt the set of new product technologies in-
troduced by rivals.

From a decision-making perspective, the question
of how quickly a firm should imitate its rivals has
been explored primarily from the perspective of first-
mover advantage (Lieberman & Montgomery, 1988).
The merit of moving first with a new product has
been extensively argued and documented (Makadok,
1998). Researchers, however, have also come to

2017 1887Giachetti, Lampel, and Li Pira

recognize that firms that move later can avoid many
of the risks that confront first movers by observing,
analyzing, and then imitating their products and
technologies (Lieberman & Montgomery, 1998;
Markides & Geroski, 2004). What is less certain is
how quickly late movers have to act if they want to
minimize risks and maximize the advantages of early
information. Studies in the competitive dynamics
and first-mover literature have suggested that, on the
whole, fast imitators—i.e., firms that imitate earlier
than others pioneering innovations—will generally
do better than firms that are slow to imitate (Lee et al.,
2000). The advantages of fast imitation are especially
strong in industries where first adopters of new
product technologies benefit from “spatial pre-
emption”; that is, the filling of product differentia-
tion niches before late adopters enter (Rao &
Rutenberg, 1979; Rindova et al., 2010). Because
spatial preemption limits the product differentiation
opportunities available to late adopters, we expect
rapid imitation of new product technologies to de-
liver higher performance for imitators that move
faster. In other words, higher average speed of imi-
tation of new product technologies offers the focal
firm more differentiation opportunities with respect
to later imitators, and is likely thereafter to lead to
higher sales volume.

The advantages of quick imitation of new product
technologies, however, are not confined to spatial
preemption. Quick imitation also has a significant
impact on consumer perception of firm reputation.
Research has shown that consumers tend to view firms
that quickly adopt new technologies as generally more
innovative (Alpert & Kamins, 1995; Carpenter &
Nakamoto, 1989; Kardes & Kalyanaram, 1992). This
judgment creates a “halo” effect that favorably skews
the evaluation of the firm’s product line, and hence
contributes to sales growth. In contrast, the product
lines of firms that are slow to adopt new technologies
(i.e., have low average speed of imitation) are judged
morenegativelybyconsumers.Thisnegativelyskewed
judgment tends to depress sales growth for slow
adopters. Therefore, in a context of multiple imitation
opportunities, firms with a high average speed of
imitation of new product technologies will be viewed
as technology leaders, and hence will benefit from
a higher reputation among customers that will en-
hance their sales performance. Thus, we predict:

Hypothesis 1b. An increase in the focal firm’s
average speed of imitation of new product
technologies will positively influence its
performance.

Scope and Average Speed of a Firm’s Imitation of
new Product Technologies and the Scope and
Average Speed of Rivals’

Imitative Actions

As noted earlier, Red Queen competition suggests
that as the number of focal firm actions increases, the
number of rival firm actions increases as well (Derfus
etal.,2008). Thatisbecausethe greaterthe focal firm’s
competitive activity, the more competitors are likely
to perceive a threat to their performance, which in
turn makes it more likely that they will respond
(Barnett & Hansen, 1996; Barnett & McKendrick,
2004). In other words, a focal firm’s increase in com-
petitive activity will present rivals with a challenge
that will increase in magnitude if the focal firm moves
ahead with new product offerings that leave rivals
with market spaces that are less and less valued by
customers. This threat will force rivals to respond
with competitive moves of their own in order to close
the gap and maintain their position.

Lieberman and Asaba (2006: 380) noted that
“rivalry-based imitation often proceeds over many
rounds, where firms repeatedly match each other’s
moves.” Generally speaking, rivalry encourages im-
itation, which in turn encourages more rivalry. The
competitive dynamics literature has suggested that
competitors that wish to maintain competitive parity
must imitate intensively (i.e., imitation scope) and
rapidly (i.e., imitation speed). This imitation effort
escalates as rivals struggle for profits and market
share. Indeed, the improved focal firm performance
derived from intense and rapid imitation of new
product technologies comes at the expense of rivals’
performance, which, in turn, may prompt rivals to
trigger aggressive imitative actions that emulate the
focal firm’s successful imitations. This gives us the
following hypotheses:

Hypothesis 2a. As the scope of the focal firm’s
imitation of new product technologies in-
creases, the scope of rivals’ imitation of new
product technologies will also increase.

Hypothesis 2b. As the average speed of the focal
firm’s imitation of new product technologies
increases, the average speed of rivals’ imitation
of new product technologies will also increase.

Scope and Average Speed of Rivals’ Imitation of
New Product Technologies and the Focal Firm’s
Performance

Various studies in the management and strategy
literature have analyzed whether and how the

1888 OctoberAcademy of Management Journal

intensity of competitive rivalry affects industry
members’ performance. A study by Young et al.
(1996) showed that increases in the number of rival
actions in a sample of software firms has a detri-
mental effect on the focal firm’s performance. Simi-
larly, Chen and Miller’s (1994) and Smith et al.’s
(1991) analyses of competitive dynamics in the air-
line industry showed that when rivals respond more
strongly to earlier moves by the focal firm, perfor-
mance of the latter will decrease. They suggested that
the more actions rivals carry out, and the greater the
speed of execution, the more the focal firm’s perfor-
mance will be damaged.

Likewise, in their analysis of Red Queen compe-
tition, Derfus et al. (2008) showed that when the focal
firm undertakes a new competitive action, both the
number and speed of rival countermoves increase,
leading to a decrease in focal firm performance.
Overall, extant studies have pointed to broader and
faster imitation by rivals as having a negative impact
on focal firm performance. This gives us the follow-
ing hypotheses:

Hypothesis 3a. With the scope of the focal firm’s
imitation of new product technologies held
constant, as the scope of rivals’ imitation of new
product technologies increases, focal firm per-
formance decreases.

Hypothesis 3b. With the average speed of the
focal firm’s imitation of new product technolo-
gies held constant, as the average speedof rivals’
imitation of new product technologies in-
creases, focal firm performance decreases.

The Moderating Effect of Product Technology
Heterogeneity in the Market

Recent studies in the strategy and technology in-
novation literature (Argyres, Bigelow & Nickerson,
2015; Giachetti & Lanzolla, 2016; Madhok et al., 2010;
Posen et al., 2013) have suggested that evolving in-
dustry characteristics, in particular changes caused
bythe introduction of newtechnologies, can affectthe
level of uncertainty in the competitive environment.
This in turn constrains the firms’ ability to learn from
rivals, reducing the effectiveness of imitation as
a competitive weapon. These findings are in line with
previous work on the industry life cycle (e.g.,
Utterback & Suarez, 1993), which has pointed out
that as industries mature they tend to transition
from high to low levels of product technology
heterogeneity—where high levels of heterogeneity

correspond to a situation in which there are more
designs contending for consumer attention, and more
product features that can be incorporated into prod-
ucts. In other words, the level of product technology
heterogeneity expresses the extent to which products
launched by all competitors are equipped with simi-
lar or different technologies. A low level of product
technology heterogeneity is the result of a “high de-
gree of design dominance,” while a high level of
product technology heterogeneity is the product of
a “low degree of design dominance.”

Since high product technology heterogeneity entails
a situation in which a clear dominant design has yet to
emerge, often because several key technologies are
vying for acceptance, firms in such an environment
have to cope with technological uncertainty when it
comes to deciding which technologies they should
install in their products (Lippman & Rumelt, 1982;
Makadok, 1998; Utterback & Suarez, 1993). One way
for firms to deal with technological uncertainty is to
observe the technologies that rivals imitated pre-
viously. However, the information obtained from
observing rivals’ imitation when technological un-
certainty is high is more noisy, and hence a less re-
liable guide for judging the merits of new product
technologies (Posen & Levinthal, 2012). In rapidly
changing competitive environments, as is the case in
Red Queen competition, technological uncertainty
can therefore slow down the learning process, con-
strain decision making, and hence adversely affect
performance. As Barkema et al. (2002: 921) pointed
out, “organizations that learn slowly from competi-
tors may find their innovation performance rapidly
deteriorating.”

3

This leads us to argue that the extent to which
a focal firm’s and rivals’ imitative actions affect the
focal firm’s performance (Hypotheses 4a and 4b, and
6a and 6b), and the extent to which the focal firm’s
imitative actions trigger rivals’ imitative actions
(Hypotheses 5a and 5b), depends on the level of
product technology heterogeneity.

Product technology heterogeneity: Focal firm’s
scope and average speed of imitation and focal
firm performance. As we noted earlier, high product
technology heterogeneity increases imitative un-
certainty. This means that focal firms are less certain

3 As also remarked by Posen and Levinthal (2012) in
their analysis of turbulent (i.e., rapidly changing) envi-
ronments, “turbulence reduces the value of efforts to gen-
erate new knowledge becausethelifespan of returns to new
knowledge is reduced in a world in which change is more
frequent” (594).

2017 1889Giachetti, Lampel, and Li Pira

about which product technologies they should imi-
tate, and which they should ignore. It also means that
the learning process for focal firms is more difficult,
since in this uncertain scenario firms need time and
resources to figure out which are the most effective
technology adoption strategies. Thus, although, in
general, we expect focal firms that are particularly
“active” when imitating new product technologies
(i.e., high imitation scope and speed) to stand a better
chance of successfully differentiating their offerings
when compared to imitating rivals that are less active,
this prediction may not hold when product technol-
ogy heterogeneity is high. When product heteroge-
neity is high firms that adopt many new product
technologies (i.e., high imitation scope), and do so
more quickly than their rivals (i.e., high imitation
speed) also run the risk of betting against the design
that will subsequently gain wide market acceptance.
Thedecisiontobetagainst afuturedominantdesignis
likely to adversely affect the performance of the focal
firm (Argyres, Bigelow, & Nickerson, 2015; Utterback
& Suarez, 1993). In contrast, low product technology
heterogeneity (i.e., high design dominance) reduces
imitation risks, largely because it is easier to evaluate
the merits of new product technologies sufficiently
early to avoid making the wrong design decisions. We
thus posit that:

Hypothesis 4a. Product technology heterogene-
ity negatively moderates the relationship be-
tween the focal firm’s scope of imitation of new
product technologies and its performance.

Hypothesis 4b. Product technology heterogeneity
negatively moderates the relationship between
the focal firm’s average speed of imitation of new
product technologies and its performance.

Product technology heterogeneity: Focal firm’s
scope and average speed of imitation and rivals’
imitation response. Various studies on organizational
learning have examined how rival firms use imitation
whentheperformanceoutcomesoflearningfromother
firms are uncertain. For example, Rhee, Kim and Han
(2006: 504) pointed out that “decision makers con-
fronting conflicting mimetic requirements and prac-
tices find it difficult to make an imitation decision
because conformity to one undermines the isomorphic
support of other elements.” Likewise, Cameron (2005)
showed that decision makers who face conflicting ex-
ternal information reduce the attention paid to such
data when updating their private information, and are
thenlikelytomakestrategicdecisionsthatdeviatefrom
industry norms. In essence, evidence has suggested

that obstacles to processing observed information—
caused by heterogeneous information—reduce imita-
tion (Gaba & Terlaak, 2013).

When product technology heterogeneity is high,
rivals confront markets in which many product
configurations compete. Under these conditions it is
unclear which of these configurations will prevail
and which will fail. Nor can rivals assume that the
entire set of actions by the first imitators conveys
information that is necessarily reliable and useful for
their imitation decisions. Their best course of action
is to keep their strategic options more open, and
imitate with greater caution, in terms of both scope
and speed. The aim of rivals at this point is to reduce
the risk of betting too early on product features that
may not become part of the future dominant design.
This means that rivals, having observed the focal
firm’s imitative actions, will imitate a limited num-
ber of technologies, and do so at lower speed. At the
industry level, this behavior leads to reduced prob-
ability of overreaction to new product technologies
that are introduced by earlier movers.

Generally speaking, therefore, the technological
uncertainty triggered by high product technology het-
erogeneity mitigates the pressure for imitative band-
wagons(Abrahamson& Rosenkopf,1993).4 Incontrast,
when there is low product technology heterogeneity,
i.e., high degree of design dominance, there is also
lower technological uncertainty because the market
features fewer product configurations. Rivals can
therefore infer more accurately the moves that focal
firms are likely to make, and hence calculate with
greater certainty the consequences of their moves. This
in turn encourages rivals to pursue imitative actions
more aggressively (i.e., higher imitation scope and
speed). This gives us the following hypotheses.

Hypothesis 5a. Product technology heterogene-
ity negatively moderates the relationship be-
tween the scope of the focal firm’s imitation of
new product technologies and the rivals’ scope
of imitation of new product technologies.

Hypothesis 5b. Product technology heterogene-
ity negatively moderates the relationship be-
tween the average speed of the focal firm’s
imitation of new product technologies and the
rivals’ average speed of imitation of new product
technologies.

4 In a similar vein, LiCalzi and Marchiori (2013) argued
that in a dynamic environment it is more effective to focus
on a relatively narrow set of strategic actions in order to
track and adapt to environmental shocks accurately.

1890 OctoberAcademy of Management Journal

Product technology heterogeneity: Rivals’ scope
and average speed of imitation and focal firm
performance. When deriving Hypothesis 5, we ar-
gued that high product technology heterogeneity
reduces rivals’ propensity to respond to the focal
firm with imitation. This is because, given the high
technological uncertainty, rivals are likely to keep
their options more open, and follow the focal firm’s
actions only if they prove to be successful. In fact, by
imitating first, the focal firm runs the risk of betting
on a design that will not become dominant (Hy-
pothesis 4), whereas rivals, by imitating later, avoid
wasting resources by imitating only those new
technologies (previously adopted by the focal firm)
that have demonstrated greater acceptance by con-
sumers. We can regard these rival firms as “second-
mover” imitators that derive their advantage from the
technological uncertainty of the market (Lieberman &
Montgomery, 1998). To put this in perspective, rivals’
imitative decisions of new product technologies (in
terms of scope and speed) will benefit from high
technological uncertainty at the expense of the focal
firm’s performance because they are able to adjust
their actions after observing the focal firm’s earlier
moves. This leads to the following hypotheses:

Hypothesis 6a. Product technology heterogene-
ity negatively moderates the relationship be-
tween the scope of the rivals’ imitation of new
product technologies and the focal firm’s
performance.

Hypothesis 6b. Product technology heterogene-
ity negatively moderates the relationship be-
tween the average speed of the rivals’ imitation
of new product technologies and the focal firm’s
performance.

Figure 1 depicts our research model, showing
the hypothesized relationships as described
above.

METHOD

Sample and Setting

We test the proposed hypotheses in the specific
context of the U.K. mobile phone industry. Our
sample includes handset vendors that were operat-
ing in the U.K. mobile phone industry from 1997 to
2008. During this period, 48 new product technolo-
gies were installed in 566 new mobile phones in-
troduced and sold by the following firms: Nokia,
Motorola, Samsung, LG, Ericsson, Sony, Sony-
Ericsson, Siemens, Philips, Panasonic, Sagem, NEC,

and Alcatel. These firms constituted almost the entire
U.K. mobile handset industry. Mobile phones can
be distinguished into two categories: (a) “regular
phones,” or “feature phones,” offering mainly basic
phone and multimedia functionalities, and (b)
“smartphones,” namely handsets equipped with ad-
vanced operating systems offering PC-like capabil-
ities that are more expensive than regular phones and
targeted at the high-end market. Smartphones con-
stitute most of the U.K. market today, but were a small
niche during the period under study. To maintain
consistency, we decided to exclude smartphone de-
vices from our sample. Information about product
innovations introduced by the 13 mobile phone ven-
dors in the U.K. market were collected from the spe-
cialist industry magazines What Mobile, What
CellPhone, and Total Mobile. We selected only
producttechnologiesthatwereexplicitlyreviewedby
these magazines over our study period.

We believe that there are several reasons why the
U.K. mobile phone industry over the 1997–2008 time
period is a particularly suitable setting to test our
hypotheses about Red Queen competitive imitation.
First, the mobile phone industry, especially in de-
veloped countries such as the U.K., has often been
described as a fast-changing environment charac-
terized by rapid new product technology in-
troduction and quick technological obsolescence
(Mintel International Group Limited, 1997–2008), all
theoretical factors that underline the pressure that
leads firms to aggressively adopt new technologies in
order to remain competitive.

Second, our observation period covers various
stages of the industry’s evolution. From the mid-
1990s to the end of the 2000s, the mobile phone
diffusion rate (i.e., the number of handsets per 100
habitants) grew from about 10% to a saturation level
(over 100%), with the growth rate of diffusion par-
ticularly high during the second half of the 1990s,
and gradually diminishing over the 2000s.5 More-
over, the progressive transition of handsets in the
U.K. from niche to mass- market products encour-
aged competitors to launch their most advanced
models and technologies in the market, making the
competitive environment particularly challenging.
These factors indicate that over the 12-year period
analyzed, the industry passed from the growth to the
maturity stage of its life cycle. Because our data
covers both growth and maturity, we are able to

5 Data about mobile phone diffusion in the U.K. market
were collected from Ofcom, the U.K. telecoms regulatory
body.

2017 1891Giachetti, Lampel, and Li Pira

examine changes in the competitive interactions and
learning processes that may occur as the technology
environment evolves over time (Baum et al., 2000).
This is in line with Derfus et al.’s (2008) recom-
mendation that research on Red Queen effects
should study empirical settings covering both early
and late stages of the industry’s evolution.

Third, mobile phone vendors in our sample are
very large companies that extensively advertise their
product innovations in a wide variety of media and
marketing channels. This means that competitive
actions related to product innovations are highly
visible—which is an important condition to assume
that imitative actions in the U.K. mobile phone in-
dustry are taken deliberately.

Fourth, the information we gathered from several
secondary sources indicated that, at least at the Eu-
ropean level, new product technologies in the mo-
bile phone industry were introduced in more or less
the same year across all European countries.6 This
makes the U.K. a representative sample of the Euro-
pean market.

Fifth, smartphone devices were a small market
category prior to the introduction of Apple’s iPhone
and its operating system iOS in mid-2007, and the
launch of Google’s Android operating system in
2008. The introduction of these product innova-
tions triggered the rapid market decline of mobile
phones that did not use advanced operating systems
(Giachetti & Marchi, 2017). To ensure consistency in
our analysis, we decided to consider only mobile
phone technologies introduced before 2008.

New Product Technologies, Technological
Systems, and Imitation

Our study focuses on drivers and performance
outcomes of new product technology imitations by
U.K. mobile phone companies. We define a product
technology as any hardware or software that allows
the handset to perform a certain function. We assume
that a “new product technology imitation” occurs
after a new product technology is introduced for the
first time in the U.K. market by a “technology pio-
neer,” or “innovator.” A firm is coded as an “imita-
tor” when it adopts for the first time in one of its new
handset models the technology previously in-
troduced by the pioneer. In our analysis, we want to
consider only the imitation of new product technol-
ogies, namely those technologies only recently in-
troduced and not widely adopted by competitors.
We consider a product technology to be widely
adopted by industry members if it has been installed
in more than 50% of all products launched in the

FIGURE

1

Research Model

Product technology
heterogeneity in the
market

Rivals’ scope and
average speed of
imitation of new
product technologies

Focal firm’s scope and
average speed of
imitation of new
product technologies

Focal firm
performance

H5a/b (-)

H2a/b (+)
H4a/b (-)

H6a/b (-)

H3a/b (-)

Time t Time t + 1 Time t + 2

Temporal sequence of actions and performance outcomes

H1a/b (+)

6 The secondary sources from which we gathered in-
formation about the timing of new product technologies
introduction were: (a) the FACTIVA database, which
searches thousands of media sources at the worldwide
level; (b) the mobile phone vendors’ annual reports and
newsletters; (c) various online catalogs for handsets, such
as the GSMArena website (http://www.gsmarena.com);
(d) books, newspapers, press releases, and business
publications.

1892 OctoberAcademy of Management Journal

http://www.gsmarena.com

market. Above this level of adoption, imitation of the
technology is no longer motivated by direct rivalry,
but by recognition that consumers now see these
features as intrinsic to the basic design and thus will
not purchase handsets that lack these features. In
total, we observed about 600 imitative actions by
firms that fit this criterion.

Since technologies may evolve over time, we fol-
low the suggestion of Giachetti and Dagnino (2017)
and analyze new product technology imitation by
considering both the first version of a technology
introduced in the market, and successive improve-
ments. A list and description of the sampled product
technologies is presented in Appendix A.

It is important to bear in mind that handsets com-
pete by offering consumers functionalities that are
made possible by product technologies. In some in-
stances, similar functionalities may be offered by
different product technologies. Following the work
on complex systems of Murmann and Frenken
(2006), we define a “technological system” as a
group of technologies that allow the product to per-
form functions of a certain type. For example, in
mobile phones infrared, Bluetooth, and USB ports
are technologies that enable connectivity between
devices, and thus belong to the same technological
system. We grouped the 48 technologies into seven
technological systems: networking, high-speed data
transfer, phone call, connectivity, messaging, dis-
play, and technological convergence (see AppendixA,
Table A1).

As can be expected, we found innovation and
imitation in all the technologies in our sample.
However, when we examined the frequency of both,
we also found that over the analyzed time period, the
average number of new product technologies in-
troduced every year—i.e., innovations—was much
lower than the average number of imitations (see
Figure A1 in Appendix A). This finding corroborates
what has been noted by previous studies: imitation is
far more pervasive than innovation. Thus, firms may
forgo the risks of innovative moves, but they cannot
avoid imitation without suffering erosion of their
market position (Lee et al., 2000; Levitt, 1966). It is
also interesting to note that the average number of
imitations rapidly increased until 2003, but started
decreasing from 2004, and the average number of
innovations was relatively high until 2003, declined
in 2004, and then leveled off from then on. The main
reason for this decline of innovations and imitations
was the shift in the locus of technological innovation
to smartphone devices. The regular phone market at
this point in time entered a period of greater emphasis

on price competition, with consequent decline in the
rates of innovation and imitation.

Measures

Dependent and independent variables. Depend-
ing on the relationship modeled in the proposed Red
Queen competitive imitation cycle (Figure 1), we rely
on a different set of dependent and independent var-
iables. We assume that the focal firm’s imitative ac-
tions at a certain time, t, trigger rivals’ response in the
following time, t 1 1, and both the focal firm’s imita-
tive action and rivals’ response will affect the focal
firm’s performance at time t 1 2, as illustrated in
Figure 1. Setting dependent and independent vari-
ables in a logical temporal sequence is important to
make realistic assumptions about the fact that ac-
tions and reactions are deliberate, and take some
time before having an effect on performance.7 De-
pendent and independent variables are described as
follows.

Consistent with the extant literature (Derfus et al.,
2008), we defined and measured the scope of a firm’s
imitation as the total number of new product tech-
nologies (belonging to a specific technological sys-
tem) imitated by the focal firm within the year t.

We measured the average speed of focal firm’s
imitation as the average time it takes for the focal
firm to imitate new product technologies related to
a specific technological system. Essentially, we
wanted to capture the speed of imitation of those
new product technologies used to operationalize
the imitation scope. To do this, we first computed
the time to imitation, in months, per each of the
technologies imitated by the firm in year t. Second,
we normalized this latter value by dividing it by the
maximum imitation time for that technology in the
sample, so as to transform the variable from count to
ratio. Third, we computed the mean of the firm’s
imitation timing of technologies belonging to the
technological system i (avtimei,t). We finally oper-
ationalized the average speed of the focal firm’s
imitation (ASi,t), as in Equation 1. The resulting
measure ranges from 0 to 1; the greater its value
(i.e., closer to 1) the higher the focal firm’s imitation
speed.

7 Since the variable rivals’ imitative response was com-
puted at time t 1 1 and the variable focal firm performance
was computed at time t 1 2, our empirical analysis cap-
tures imitative actions between 1997 and 2007, and firm
performance from 1999 and 2008.

2017 1893Giachetti, Lampel, and Li Pira

ASi,t 5 1 2
�
avtimei,t

�
(1)

It is worth noting at this point that a higher average
speed of imitation does not entail higher imitation
scope. In fact, two focal firms may have the same
score for average imitation speed but imitate a dif-
ferent number of technologies. Moreover, if one firm
increases the number of technologies imitated from
one year to another (i.e., wider scope), this might
result in either higher or lower average speed with
respect to the previous year (e.g., “lower speed” if the
firm imitates a “higher number” of technologies, but
“more slowly”).

We operationalized the scope of a rivals’ imitation
by subtracting the total number of imitations realized
by the focal firm at time t 1 1 from the total number of
imitative actions taken by all competitors at the same
time, t 1 1 (in a focal technological system). In this
way, we accounted only for those imitative actions
subsequent to the focal firm’s imitative actions.

As in other competitive dynamics research
(e.g., Ferrier et al., 1999; Young et al., 1996), we used
rivals’ imitation speed as a measure of the average
length of time it took rivals to act after a new product
technology was introduced. Following the procedure
outlined by Derfus et al. (2008), we calculated this
measure by taking the mean of the average speed of
imitation of all of the focal firm’s rivals at a certain
time, t 1 1. The resulting measure ranges from 0 to 1,
with the higher imitation speed for values closer to 1.

Focal firm performance was operationalized us-
ing the number of handsets sold on a yearly basis
(i.e., sales performance) in the U.K. This measure of
firm performance has been widely used by mobile
phone industry specialists such as Gartner Data-
quest and Mintel International Group Limited. Data
on handsets sold per vendor were collected from
Mintel International Group Limited (1997–2008),
Euromonitor International (2003–2008), and firms’
archival data.

We operationalized the measure of product tech-
nology heterogeneity using the Shannon entropy
index (Shannon, 1948). This entropy measure is
suitable for our research setting because it captures
the extent to which products differ in terms of tech-
nologies that belong to a given technological system.
A uniform distribution of the type of technologies
products are equipped with reflects a situation in
which firms produce a wide variety of designs, while
a skewed distribution represents a situation in which
there are minor differences between firms’ choice of
design. As such, theindex can be used as an indicator
of technological heterogeneity (Frenken, Saviotti, &

Trommetter, 1999), a situation in which products
offered by industry rivals widely differ in terms of the
technologies they are equipped with. The Shannon
entropy value of a technological system is given by
Equation 2:

Hi,t 5 2 +
S

k 5 1
ln
�
pk,t

�
3 pk,t (2)

Where Hi,t is the level of product technology het-
erogeneity within technological system i at year t, pk,t
is the percentage of products (introduced in year t)
equipped with technology k (therefore 0 # pk # 1),
and S is the number of technologies introduced and
related to technological system i.

The Shannon entropy index (Hi,t) is equal to zero
when all products introduced at time t in the market
are equipped with the same set of technologies re-
lated to technological system i. This means that there
is a dominant design in terms of the set of technolo-
gies related to i. In this extreme case, pk,t would be
equal to 1, which implies that the entropy of the
product population equals zero:

Hi,t 5 2lnð1Þ 3 1 5 0 (3)
Entropy is positive otherwise, and the larger its

value, the larger the variety in the population. Spe-
cifically, the larger the value of Hi,t, (a) the higher the
number of technologies in the technological system,
and (b) the lower the diffusion of these technologies
among existing products.

Control variables. We also included various
control variables (those related to the focal firm and
at the industry level are computed at year t, those
related to rivals are computed at year t 1 1), poten-
tially affecting all firms’ action and performance:

Although we are analyzing competitive dynamics
that are triggered by imitative efforts, we had to
control for imitation that occurs as a response to in-
novations introduced into the technological system,
or what we call innovation scope. This is in line with
first-mover advantage literature, which has sug-
gested that innovators’ monopoly profits will attract
imitative entrants (Lieberman & Montgomery, 1988;
Markides & Geroski, 2004). This variable was mea-
sured as a count of new product technologies in-
troduced by the focal firm in year t.

Similar to how we measured the focal firm’s in-
novation scope, we measured rivals’ innovation
scope as a count of the new product technologies
introduced by rivals in the year t 1 1.

Studies of the Red Queen effect have argued that
a firm’s relative size can influence its performance,

1894 OctoberAcademy of Management Journal

and rivals’ responseto its actions (Derfus et al., 2008).
Relative market position was measured with a
dummy variable that set the value as 1 if the level of
sales of the firm in the year t was above the industry
median, and 0 otherwise.

Mobile phone vendors may follow different strat-
egies depending on the time of year in which they
introduce the largest number of new product models.
We controlled for this strategic decision with a set of
dummy variables that equaled 1 during the quarter
when the firm introduced the largest number of new
product models during the year t, and 0 otherwise.

Researchinindustrialorganizationandstrategyhas
shown that industry concentration can influence the
intensity of competition (Derfus et al., 2008). In an
industrywith highbarriersto entry, suchasthemobile
phone industry, a higher level of industry concentra-
tion usually results in a lower level of competition
intensity, because rivals with the largest market share
are more likely to collude on their marketing strategies
(Waldman&Jensen,2012;Wiggins&Ruefli,2005).We
therefore controlled for industry concentration by us-
ing the cumulative market share of the four largest
U.K. handset vendors as a measure.

A three-year standard deviation of the U.K. gross
domestic product (GDP volatility) was used to ac-
count for the country’s macroeconomic uncertainty
(Haddow, Hare, Hooley, & Shakir, 2013).

RESULTS

Hypotheses Testing

Table 1 reports the variables’ descriptive statistics,
while Tables 2–3 report results of the regression
analysis. We tested the hypotheses with three re-
gression models: (1) a robust fixed-effects regression
when the dependent variable was the focal firm
performance (Table 2); (2) a robust fixed-effects re-
gression when the dependent variable was rivals’
average speed of imitation (Table 3); (3) a robust
fixed-effects Poisson regression when the dependent
variable was rivals’ imitation scope, a count-type
variable (Cameron & Trivedi, 2009) (Table 3). A
Hausman test suggested that the use of fixed-effects
was preferable over random-effects. Since not all
technologies were adopted by all sampled firms, and
not all firms were active in the U.K. market over the
entire time period analyzed, we ended up with a
566-observation unbalanced panel.

Models 1–3 in Table 2 report the results for re-
gressions relating focal firm imitation scope and
speed, and rivals imitation scope and speed, to firm

performance(Hypotheses1,3,4,and6).Theregression
results that examine the impact of focal firm imitation
scope and speed on rivals’ imitation scope and speed,
respectively, are presented in Table 3, Models 4–9
(Hypotheses 2 and 5). We calculated variance inflation
factors (VIFs) to determine whether there was multi-
collinearity in the analyses. The average VIF scores
were all below 1.4, and no individual VIF was greater
than 2.08, thereby all were lower than the recom-
mended threshold of 10 (Chatterjee & Hadi, 2006).8

Before we turn to a discussion of the coefficients of
independent variables and moderators related to the
presented hypotheses, we briefly examine the co-
efficients of the control variables in the full Models 3
(Table 2), 6, and 9 (Table 3). We found the impact of
innovation scope on focal firm performance, as
shown in Model 3 (Table 2), in terms of both focal
firm innovation scope (b 5 0.00, p . .1) and rivals’
innovation scope (b 5 20.00, p . .1), not to be sig-
nificant. With regard to the impact of innovation
scope on imitative actions, as shown in Models 6 and
9 (Table 3), we found that the only significant re-
lationship is that between rivals’ innovation scope
and rivals’ imitation scope (Model 6: b 5 0.26,
p , .01), showing that rivals that innovate more are
also those that imitate more.9 We also found that the
control variable relative market position has a sig-
nificant effect only on focal firm performance as
shown in Model 3 (b 5 0.27, p , .01). As for industry-
level controls, industry concentration has a negative
and significant effect on rivals’ average speed of
imitation (Model 9: b 5 20.33, p , .01), while GDP
volatility has a positive effect on focal firm perfor-
mance (Model 3: b 5 0.05, p , .01) and a negative

8 As can be observed in the correlation matrix presented
in Table 1, the greatest correlation coefficient is that be-
tween focal firm’s imitation scope and speed (r 5 0.63;
p , .01), two key independent variables in our regression
model. In the regression models, the maximum VIFs for
these two variables were 2.08 and 1.79, respectively.

9 The positive association between firm’s imitation
scope and innovation scope can also be observed in the
correlation matrix presented in Table 1; the correlation
coefficients between rivals’ imitation scope and rivals’
innovation scope (r 5 0.35, p , .01) and between focal
firm’s imitation scope and focal firm’s innovation scope
(r 5 0.08, p , .1) are both positive and significant. We
believe the explanation for this is that firms with greater
resources and the capabilities needed to imitate several
technologies also have greater resources and capabilities to
introduce several technologies that are new to the market,
and vice versa.

2017 1895Giachetti, Lampel, and Li Pira

T
A
B
L
E
1

D
es
cr
ip
ti
v
e
S
t

a
ti

st
ic
s

M
ea

n
S
D

1
2

3

4

5

6

7

8

9

1
0

1
1

1

F
o
ca

l
fi
rm

’s
p

er
fo

rm

an
ce

(t
1

2)
a

1
7
9
7
.2
1

2
2
7
5
.3
1

1
.0
0

2
F
o
ca
l
fi
rm

’s
im

it
at
i

o
n

sc
o
p
e

(t
)

0
.5
1

0
.8
2

0
.0
8
1

1
.0
0

3
F
o
ca
l
fi
rm

’s
av

er
ag

e
sp

ee
d
o
f

im

it
at

io
n

(t
)

0
.1
5

0
.2
6

0
.1
7
*
*

0
.6
3
*
*

1
.0
0

4
R
iv
al
s’
im

it
at
io
n
sc
o
p
e

(t
1

1
)

4
.7
2

4
.4
1

2

0
.0
6

0
.3
0

*
*

0
.2
0
*
*

1
.0
0

5
R
iv
al
s’
av

er
ag
e
sp
ee
d
o
f

im
it

at
io
n
(t
1

1
)

0
.3
5

0
.2
4

2
0
.1
7
*
*

2

0
.0
7

1

0
.0
4

0
.0
9
*

1
.0
0

6
P

ro
d

u
ct

te
ch

n
o

lo
gy

h
et
er
o
ge

n
ei
ty

(t
)

1
.4
6

0
.8
7

0
.1
4
*
*

0
.4
4
*
*

0
.2
0
*
*

0
.4
1
*
*

2
0
.2
9
*
*

1
.0
0

7
F
o
ca
l
fi
rm

’s
in
n
o
v
at
io
n

sc
o
p
e
(t
)

0
.0
7
0
.3
0
0
.0
8
1
0
.0
8
1
0
.0
9
*

0
.1
1
*
*

0
.0
7
0
.0
6
1
.0
0

8
R
iv
al
s’
in
n
o
v
at
io
n
sc
o
p
e

(t
1
1
)

0
.5
7

0
.8
0

2

0
.1
0

*

2

0
.0
5

2

0
.0
3

0
.3
5
*
*

0
.1
9
*
*

2
0
.0
9
*

2
0
.0
0

1
.0
0

9
F
o
ca
l
fi
rm

’s
re
la
ti
v
e

m
ar
k
et

p
o
si
ti
o
n
(t
)

0
.5
0

0
.5
0

0
.6
1
*
*

0
.0
7
0
.1
7
*
*

2

0
.0
2

2
0
.0
5

0
.0
3
0
.0
5

2
0
.0
3

1
.0
0
1
0

I

n
d

u
st
ry

co
n
ce
n
tr
at
io
n
(t
)

0
.7
8

0
.0
5
0
.0
8
1
0
.0
2

2
0
.0
4

0
.0
1

2
0
.2
5
*
*

0
.1
0
*

2
0
.0
1

2
0
.0
0
0
.0
8
1
1
.0
0
1
1

G
D
P
v
o
la
ti
li
ty

(t
)b

1
0
5
7
3
.8
5

2
6
8
3
.7
7

2
0
.0
8
1

2
0
.2
2
*
*

2
0
.0
9
*

2

0
.1
2
*
*

0
.1
2
*
*

2
0
.3
5
*
*

2
0
.0
2

0
.0
6
2
0
.0
5

2
0
.4
9
*
*

1
.0
0

N
o
te
:n

5
5
6
6

a
U
n
it
s
so
ld

ar
e
ex

p
re
ss
ed

in
th
o
u
sa
n
d
s.

b
G
D
P
v
o
la
ti
li
ty

is
co

m
p
u

te
d

o
n
G
D
P
v
al
u
es

in
m
il
li
o
n
s
o
f
p
o
u
n
d
s.

1
p
,

0
.1
0

*
p
,

0
.0
5

*
*
p
,

0
.0
1

1896 OctoberAcademy of Management Journal

effect on rivals’ imitative actions (Model 6: b 5
20.07, p , .05; Model 9: b 5 20.23, p , .01).

We now turn our attention to the hypotheses tests.
Hypotheses 1a and 1b state that focal firm imitation
scope and average speed of imitation both have
a positive effect on its performance. As shown in
Model 3 (Table 2), while the sign and significance of

focal firm average speed of imitation is in line with
our prediction (b 5 0.07, p , .05), focal firm imita-
tion scope is significant with the opposite sign (b 5
20.08, p , .05). Therefore, Hypothesis 1b is sup-
ported while Hypothesis 1a is not.

Hypothesis 2a states that as the scope of the firm’s
imitation of new product technologies increases, the

TABLE 2
Robust Fixed-effects Regression Analysis: Focal Firm and Rivals’ Imitative Actions on the Focal Firm Performance

Model 1 Model 2 Model 3

Hypothesis

Focal firm’s

performance (t 1 2)

Focal firm’s

performance (t 1 2)
Focal firm’s
performance (t 1 2)

Constant 20.10** 20.05* 20.06**
(–5.23) (–2.47) (–2.76)

Independent variables
Focal firm’s imitation scope (t) 1a 20.041 20.08*

(–1.86) (–2.43)
Focal firm’s average speed of imitation (t) 1b 0.05* 0.07*

(2.00) (2.50)
Rivals’ imitation scope (t 1 1) 3a 20.05** 20.04*

(–2.76) (–2.39)
Rivals’ average speed of imitation (t 1 1) 3b 20.041 20.051

(–1.98) (–1.95)
Product technology heterogeneity (t) 0.12* 0.12**

(2.62) (2.99)
Interactions
Focal firm’s imitation scope 3 Product
technology heterogeneity

4a 0.05*
(2.03)

Focal firm’s average speed of imitation 3
Product technology heterogeneity

4b 0.00
(0.18)

Rivals’ imitation scope 3 Product
technology heterogeneity

6a 20.04*
(–2.31)

Rivals’ average speed of imitation 3
Product technology heterogeneity

6b 20.00
(–0.12)

Controls
Focal firm’s innovation scope (t) 20.00 0.00 0.00

(–0.28) (0.17) (0.16)
Rivals’ innovation scope (t 1 1) 20.04* 0.00 20.00

(–2.26) (0.08) (–0.04)
Relative market position (t) 0.30** 0.28** 0.27**

(5.04) (5.40) (5.32)
Industry concentration (t) 20.01 20.01 20.02

(–0.55) (–0.56) (–0.92)
GDP volatility (t) 0.04** 0.06** 0.05**

(2.70) (2.78) (2.65)
2nd quarter year t (largest new product
launch)

0.20** 0.15** 0.15**
(4.96) (3.84) (3.86)

3rd quarter year t (largest new product
launch)

0.12* 0.09* 0.10*
(2.61) (2.05) (2.17)

4th quarter year t (largest new product
launch)

0.03 0.01 0.01
(0.69) (0.15) (0.36)

n 566 566 566
Within R-squared 0.24 0.30 0.31

Notes: Estimates are based on standardized variables; t-statistics in parentheses.

1p , 0.10
*p , 0.05

**p , 0.01

2017 1897Giachetti, Lampel, and Li Pira

scope of rivals’ imitation of new product technolo-
gies will also increase. Hypothesis 2b states that as
the average speed of the firm’s imitation of new

product technologies increases, the average speed of
rivals’ imitation of new product technologies will
also increase. As can be observed from Table 3, in

TABLE 3
Robust Fixed-effects Regression Analysis: Focal Firm Imitative Actions on Rivals’ Imitative Actions

Robust fixed-effects Poisson Robust fixed effects

Model 4 Model 5 Model 6 Model 7 Model 8 Model 9

Hypothesis

Rivals’
imitation

scope(t 1 1)

Rivals’
imitation
scope(t 1 1)
Rivals’
imitation
scope(t 1 1)

Rivals’ average
speed of

imitation(t1 1)

Rivals’ average
speed of

imitation (t11)

Rivals’ average
speed of

imitation(t11)

Constant 0.25** 0.05 0.05
(4.52) (0.79) (0.88)

Independent
variables
Focal firm’s
imitation scope (t)

2a 0.081 0.15** 0.04 0.05
(1.76) (3.11) (0.87) (0.93)

Focal firm’s average
speed of imitation (t)

2b 0.06 0.02 0.02 0.02
(1.61) (0.67) (0.52) (0.42)

Product technology
heterogeneity (t)

0.23** 0.23** 20.62** 20.61**
(4.30) (4.53) (–14.78) (–14.19)

Interactions
Focal firm’s imitation
scope 3 Product
technology
heterogeneity

5a –0.09*
(–2.38)

Focal firm’s average
speed of imitation 3

Product technology
heterogeneity

5b –0.02
(–0.72)

Controls
Focal firm’s
innovation scope (t)

0.02 0.02 0.02 0.05 0.02 0.02
(0.52) (0.63) (0.64) (1.18) (0.44) (0.45)

Rivals’ innovation
scope (t 1 1)

0.18** 0.26** 0.26** 0.09* 20.03 20.02
(5.26) (6.81) (7.15) (2.03) (–0.56) (–0.54)

Relative market
position (t)

20.05 20.07 20.05 20.07 20.01 20.01
(–1.11) (–1.46) (–1.13) (–0.88) (–0.22) (–0.20)

Industry
concentration (t)

20.07 0.00 20.00 20.28** 20.33** 20.33**
(–1.40) (0.07) (–0.10) (–7.53) (–10.15) (–10.14)

GDP volatility (t) 20.17** 20.061 20.07* 20.07 20.23** 20.23**
(–6.44) (–1.94) (–2.16) (–1.45) (–5.60) (–5.60)

2nd quarter year t
(largest new
product launch)

20.03 20.12 20.14 20.32** 20.09 20.09
(–0.31) (–1.24) (–1.42) (–3.06) (–0.85) (–0.87)

3rd quarter year t
(largest new
product launch)

20.04 20.06 20.08 20.27* 20.14 20.14
(–0.39) (–0.58) (–0.81) (–2.24) (–1.30) (–1.32)

4th quarter year t
(largest new
product launch)

0.11 0.06 0.04 20.19 1 20.03 20.04
(1.25) (0.63) (0.38) (–1.71) (–0.32) (–0.34)

n 566 566 566 566 566 566
Within R-squared 0.09 0.23 0.23
Wald x2 86.25 113.00 132.65

Notes: Estimates are based on standardized variables; in parentheses are reported t-statistics for robust fixed-effects and z-statistics for robust
fixed-effects Poisson; coefficients in bold are those related to the tested hypotheses.

1p , 0.10
*p , 0.05
**p , 0.01

1898 OctoberAcademy of Management Journal

Model 6 the relationship between the scope of the
focal firm’s imitation and the scope of rivals’ imita-
tion is positive and significant (b 5 0.15, p , .01),
while in Model 9 the relationship between the av-
erage speed of the focal firm’s imitation and the av-
erage speed of rivals’ imitation is positive but not
significant (b 5 0.02, p . .1). Therefore, Hypothesis
2a is supported, while Hypothesis 2b is not.

Hypothesis 3a states that with the scope of the fo-
cal firm’s imitation of new product technologies held
constant, as the scope of rivals’ imitation of new
product technologies increases, focal firm perfor-
mance decreases. Hypothesis 3b leads us to expect
that holding the average speed of the focal firm’s
imitation of new product technologies constant, we
can observe decreasing focal firm performance as the
average speed of rivals’ imitation of new product
technologies increases. As seen in Model 3 (Table 2),
the coefficient of scope and average speed of rivals’
imitation are both negative and significant (b 5
20.04, p , .05; b 5 20.05, p , .1), thus supporting
both Hypotheses 3a and 3b.

Hypothesis 4a states that product technology het-
erogeneity negatively moderates the relationship
between the focal firm’s scope of imitation of new
product technologies and its performance. Hypoth-
esis 4b states that product technology heterogeneity
negatively moderates the relationship between the
focal firm’s average speed of imitation of new prod-
uct technologies and its performance. As shown in
Model 3, the coefficient of the interaction between
focal firm imitation scope and product technology
heterogeneity is positive and significant (b 5 0.05,
p , .05), and the coefficient ofthe interaction between
focal firm imitation speed and product technology
heterogeneity is not significant (b 5 0.00, p . .1).
Hypotheses 4a and 4b are thus not supported.

Hypothesis 5a predicts that product technology
heterogeneity negatively moderates the relationship
between the scope of the firm’s imitation of new
product technologies and the rivals’ scope of imita-
tion of new product technologies. Hypothesis 5b
predicts that product technology heterogeneity neg-
atively moderates the relationship between the av-
erage speed of the firm’s imitation of new product
technologies and the rivals’ average speed of imita-
tion of new product technologies. As shown in
Model 6, the coefficient of the interaction between
the focal firm’s imitation scope and product tech-
nology heterogeneity is negative and significant, as
expected (b 5 20.09, p , .05), while in Model 9
the coefficient of the interaction between the focal
firm’s imitation speed and product technology

heterogeneity is negative but not significant (b 5
20.02, p . .1). Hypothesis 5a is thus supported while
Hypothesis 5b is not.

With Hypothesis 6a, we predict that product
technology heterogeneity negatively moderates
the relationship between the scope of the rivals’
imitation of new product technologies and focal
firm performance. With Hypothesis 6b, we predict
that product technology heterogeneity negatively
moderates the relationship between the average
speed of the rivals’ imitation of new product
technologies and focal firm performance. As
shown in Model 3, the coefficient of the interaction
between rivals’ imitation scope and product tech-
nology heterogeneity is negative and significant, as
expected (b 5 20.04, p , .05), while the coefficient
of the interaction between rivals’ imitation speed
and product technology heterogeneity is negative
but not significant (b 5 20.00, p . .1). Hypothesis
6a is thus supported, whereas Hypothesis 6b is not.

Table 4 offers a summary of the predicted hy-
potheses and those that were supported by the em-
pirical analysis. As can be observed, the Red Queen
competitive imitation cycle (Hypotheses 1–3) is
supported for at least one type of imitative action in
all time frames. In the discussion section, we will
present the plots of interaction effects and extend the
interpretation of these findings.

Robustness Tests

We tested the robustness of our findings in several
ways. First, we examined an alternative explana-
tion to the one advanced in Hypotheses 1a and 1b.
Specifically, if imitation scope rises, new product
development costs will escalate, which in turn will
lead to negative performance consequences. By the
same token, as firms increase their imitation speed
to catch up with their rivals, they have less time to
adequately assess market response, and this in turn
is likely to have negative performance conse-
quences. Under both scenarios, we should expect
an inverted U-shaped relationship between both
types of imitative action and firm performance. To
test these alternative predictions, we repeated the
regression analysis by adding the squared term of
both focal firm imitation scope and average speed of
imitation. We did not find the squared terms to be
significant.

Second, we examined the robustness of our results
in light of the fact that the dependent variables—
average speed of imitation and imitation scope—are
left–and right-censored, respectively. Average speed

2017 1899Giachetti, Lampel, and Li Pira

of imitation is left-censored because it is a ratio that
cannot be less than zero, and may take the value of
zero both when a firm has minimum average imita-
tion speed and when the firm is not imitating any
technology. Imitation scope is right-censored be-
cause the number of technological attributes avail-
able to be imitated has an upper limit. We therefore
tested the full Models 6 and 9 (Table 3) using alter-
native models that took into account the censored
nature of both dependent variables. More specifi-
cally, we repeated Models 6 and 9 using a Tobit
fixed-effects regression based on the Honoré (1992)
estimator with an absolute error loss function. This
estimator was chosen because there is no conditional
fixed-effects Tobit model, and the unconditional
fixed-effects Tobit model is biased (Honoré, 1992).
As shown in Table 5, Models 10 and 11, even with
this alternative technique, results are consistent with
those presented in Table 3.

Third, since the regression equations in Models 6
and 9 rely on the same set of independent variables,
in order to account for potential correlations of the
random error components of the two equations, we
ran Models 6 and 9 using the seemingly unrelated
regressions technique (Zellner, 1962). This method
involves estimating separate equations for rivals’
speed and scope of imitation while recognizing re-
lationships across the two actions. As shown in
Table 5, Models 12 and 13, results are consistent with
those in Models 6 and 9, with the exception of

Hypothesis 5a (which presents the expected sign, but
is not significant).

DISCUSSION

Implications

This study aims to expand our understanding of
competitive dynamics in technology-intensive in-
dustries with the lens of Red Queen competition. We
do this by bringing together relevant research from
competitive dynamics, imitation, and technology in-
novation literature. The more recent Red Queen lit-
erature has analyzed the conditions under which
competitive actions increase firm performance and
trigger rivals’ response (Derfus et al., 2008), but has
not paid sufficient attention to (a) the analysis of Red
Queencompetitivedynamicsintechnology-intensive
industries, (b) the role of different types of imitative
actions in sustaining and triggering the Red Queen
cycle, and (c) how changes in the technological en-
vironment moderate the Red Queen cycle. To address
these gaps, we have developed a model of Red Queen
competition in which the scope and speed of imita-
tion of new product technologies is the result of
competitive threats by rivals’ imitative actions. The
competitive race predicted by our theory of Red
Queen competitive imitation implies that firms
struggle to (a) learn which technologies are, and will
be, successful in the market, (b) imitate new product

TABLE 4
Predicted Hypotheses and Obtained Findings

Obtained findingsa

Hypotheses Predicted relationship
Imitation scope
(Hypotheses a)

Average imitation
speed (Hypotheses b)

1 Positive effect of focal firm’s imitative actions on its
performance

Negativeb Positive

2 Positive effect of focal firm’s imitative actions on
rivals’ imitative actions

Positive Not significant

3 Negative effect of rivals’ imitative actions on focal
firm’s performance

Negative Negative

4 Negative moderating effect of product technology
heterogeneity on the relationship between focal
firm’s imitative actions and its performance

Positive Not significant

5 Negative moderating effect of product technology
heterogeneity on the relationship between focal
firm’s imitative actions and rivals’ imitative actions

Negative Not significant

6 Negative moderating effect of product technology
heterogeneity on the relationship between rivals’
imitative actions and focal firm’s performance

Negative Not significant

a Relationships supported by the empirical analysis are in bold.
b Positive for high levels of product technology heterogeneity (Figure 2).

1900 OctoberAcademy of Management Journal

technologies to maintain competitive parity with ri-
vals, and thus (c) adapt to the evolving technological
environment. The analysis of this self-reinforcing
competitive mechanism enables us to shed light on
the positive and negative aspects of different types of
imitative action, and to clarify the relative importance
of these aspects with regard to firm performance.

Our first result shows that focal firms’ average
speed of imitation positively affects their sales per-
formance (Hypothesis 1b) while, contrary to our
prediction, focal firms’ imitation scope has a detri-
mental effect on performance (Hypotheses 1a). Re-
sults of speed are consistent with previous findings
of the competitive dynamics literature (D’Aveni,

TABLE 5
Tobit Fixed-effects Regression and Seemingly Unrelated Regression Analysis: Focal Firm Imitative Actions on Rivals’

Imitative Actions

Tobit Fixed effects Seemingly Unrelated Regressiona

Model 10 Model 11 Model 12 Model 13

Rivals’ imitation
scope (t 1 1)

Rivals’ average speed of
imitation (t 1 1)

Rivals’ imitation
scope (t 1 1)
Rivals’ average speed of
imitation (t 1 1)

Constant 20.13 20.31
(–0.47) (–1.05)

Independent variables
Focal firm’s imitation scope (t) H2a 0.38* 0.14 0.15* 0.05

(1.96) (0.98) (2.54) (0.94)
Focal firm’s average speed of
imitation (t)

H2b 0.06 –0.15 0.04 0.02
(0.54) (–1.18) (0.80) (0.36)

Product technology
heterogeneity (t)

0.78** 21.02** 0.33** 20.61**
(4.35) (–6.55) (5.35) (–9.58)

Interactions
Focal firm’s imitation scope 3
Product technology
heterogeneity

H5a –0.25* –0.05
(–2.27) (–1.04)

Focal firm’s average speed of
imitation 3 Product
technology heterogeneity

H5b –0.06 –0.03
(–0.89) (–0.72)

Controls
Focal firm’s innovation scope (t) 0.01 20.03 0.02 0.02

(0.08) (–0.33) (0.58) (0.57)
Rivals’ innovation scope (t 1 1) 0.61** 20.02 0.32** 20.02

(8.37) (–0.21) (8.44) (–0.64)
Relative market position (t) 20.191 0.04 20.09 20.01

(–1.92) (0.38) (–1.44) (–0.20)
Industry concentration (t) 20.181 20.57** 20.02 20.33**

(–1.76) (–3.96) (–0.50) (–7.32)
GDP volatility (t) 20.07 20.43** 20.05 20.23**

(–0.73) (–3.52) (–1.19) (–5.60)
2nd quarter year t (largest new
product launch)

20.53** 20.04 20.13 20.09
(–2.86) (–0.23) (–1.39) (–0.98)

3rd quarter year t (largest new
product launch)

0.09 20.19 20.05 20.14
(0.35) (–1.07) (–0.49) (–1.42)

4th quarter year t (largest new
product launch)

0.24 20.321 20.05 20.04
(1.05) (–1.66) (–0.49) (–0.35)

n 566 566 566 566
R-squared 0.42 0.37
x
2 220.05 146.93 417.94 331.59

Notes: Estimates are based on standardized variables; z-statistics in parentheses; coefficients in bold are those related to the tested
hypotheses.

aFirm dummies were included but not reported.
1p , 0.10
*p , 0.05

**p , 0.01

2017 1901Giachetti, Lampel, and Li Pira

1994; Lee et al., 2000). By contrast, the negative effect
of scope on focal firm performance is apparently
counterintuitive. We will consider this result again
later in this section, since the imitation scope–
performance relationship turns out to be positive
when considering the moderating effect of product
technology heterogeneity. In line with the Red
Queen argument, we also found that focal firm imi-
tation scope triggers rivals’ imitation scope (Hy-
pothesis 2a), but focal firm speed of imitation does
not trigger rivals’ rapid imitation (Hypothesis 2b).
There are two possible interpretations of these re-
sults: (a) it is possible that rivals perceive scope as
more of a threat to their competitive positions com-
pared to speed, and thus are more likely to invest
resources matching scope rather than speed; or
(b) rivals may not be able to move as quickly as the
focal firms that were the earliest, if not the first, to
make the imitative moves. Either way, whether rivals
choose to focus resources on scope over speed, or
cannot marshal the resources to respond quickly, ri-
vals definitely respond to scope moves, implying that
scope is an important strategic issue in technology-
intensive industries. These results contrast in part
with those studies in the competitive dynamics liter-
ature that have described response speed as the main
strategic issue firms focus resources on when coun-
termoveing against rivals (Derfus et al., 2008;
Markides & Geroski, 2004). Finally, to close the Red
Queen cycle, we found that rivals’ imitation scope
and speed have a negative effect on the focal firm’s
performance (Hypotheses 3a and 3b).

To illustrate our results, it may be useful to give an
example. Nokia’s pioneering of digital technologies
such as infrared, games, an email client, and WAP
(Wireless Application Protocol) during the 1990s
elicited various reactions from rivals. Siemens was
among the first to imitate (high imitation speed) all of
the technologies mentioned earlier (high imitation
scope). This reinforced Siemens’ product portfolio
competitiveness, and increased its sales perfor-
mance relative to slower imitators such as NEC,
Philips, and Sagem. Nevertheless, Siemens enjoyed
a temporary competitive advantage that lasted until
Nokia’s innovations were adopted by other handset
vendors. Subsequently, at the beginning of the
2000s, some vendors pioneered new product tech-
nologies, such as Bluetooth, MMS (Multimedia
Messaging Service) and photo camera, and a new
series of imitative actions commenced, with firms
such as Sony-Ericsson and Samsung installing this
set of features in their new lines of phones more
quickly compared to Siemens. Although in the first

time period (i.e., during the 1990s) Siemens was
able to match the scope and speed requirements,
and in turn enjoyed a temporary competitive ad-
vantage, in the second time period (i.e., the begin-
ning of the 2000s) it did not possess the imitative
capabilities to stay aligned with rivals, and strug-
gled to catch up. To paraphrase Lewis Carroll
(1960), Siemens realized that although it was run-
ning as fast as it could, it was not getting anywhere
relative to its rivals. Interestingly, the escalating
pressure to imitate in order to retain market position
not only increased “competitive imitation” among
handset vendors, but also accelerated the techno-
logical evolution of the industry. In fact, looking
back it is remarkable how quickly the industry
moved in a few years from basic handsets capable of
providing only phone calls in the mid-1990s, to
multi-tasking devices that integrate nearly all types
of portable technologies (Figure A1).

In order to get a clearer picture of the boundaries of
Red Queen competition in a technology-intensive
industry, we also examined the extent to which
Red Queen evolution may depend upon a specific
industry condition—in our case, the level of product
technology heterogeneity in the market. We found
product technology heterogeneity to have a signifi-
cant moderating effect in all time frames of the pro-
posed Red Queen competitive imitation cycle, for at
least one type of imitative action (Table 4). First,
contrary to our prediction in Hypothesis 4a, our re-
sults indicate that product technology heterogeneity
significantly and positively moderates the effect of
focal firm imitation scope on focal firm performance.
This result, combined with the negative direct effect
of firm imitation scope on its performance, is repre-
sented in Figure 2. More specifically, when we plot
the data of the significant interaction (i.e., scope of
focal firm imitation 3 product technology hetero-
geneity), we observe that: (a) the effect of the focal
firm’s imitation scope on its performance is positive
for high levels of product technology heterogeneity,
while it is negative for low levels of product tech-
nology heterogeneity, and (b) performance gains
from the focal firm’s imitation scope are maximized
when this scope is large, and product technology
heterogeneity is high. The overall picture shows that
imitation scope may indeed have a positive effect on
firm performance, as predicted in Hypothesis 1a, but
this occurs only for high levels of product technology
heterogeneity. Ex post, an explanation for this result
could be that when product technology heterogene-
ity is high, focal firms have to imitate as many new
technologies as they can in order to increase the

1902 OctoberAcademy of Management Journal

probability of launching new product models that
converge with the product configuration that will
become dominant.

Moreover, as predicted in our theory, we found
that product technology heterogeneity negatively
moderates the relationship between the focal firm’s
imitation scope and rivals’ imitation scope. This is
because when product technology heterogeneity is
high, the focal firm’s and rivals’ learning process
is constrained. Rivals that react to the focal firm’s
moves are more likely to be conservative when it
comes to the number of new technologies imitated,
preferring to wait until the technological uncertainty
decreases. Overall, these findings are consistent with
observations by other studies in the Red Queen lit-
erature, namely that learning from competitive ex-
perience will be less effective if firms encounter
a series of environmental shocks that render their
learning capability obsolete (Barkema et al., 2002;
Derfus et al., 2008). Bearing in mind that product
technology heterogeneity changes, which in turn
influences the pace of technological change, we be-
lieve that our results also contribute to research on
how technological changes in technology-intensive
industries may influence the way firms compete
(Agarwal, Sarkar, & Echambadi, 2002; Utterback &
Suarez, 1993), as well as their ability to preserve their
performance vis-à-vis rivals (Bayus & Agarwal,
2007).

In line with our predictions, we also found that
product technology heterogeneity negatively mod-
erates the effect of rivals’ imitation scope on the

focal firm’s performance. When products differ
greatly in terms of the technologies they in-
corporate, rivals’ imitative response to focal firm
actions will disproportionately decrease the focal
firm’s performance. The main reason for this, as we
see it, is that rivals have an imitative advantage
when the focal firm confronts greater uncertainty
about the performance of new technologies. Rivals
can observe the performance outcomes of the focal
firm’s imitative action and then imitate (in the fol-
lowing period) only new technologies that have
demonstrated greater acceptance by consumers. In
this way, rivals strengthen their competitive posi-
tion with respect to the focal firm by investing only
in value-enhancing technologies. Plotting the data
from Model 3, we graphically represent the form of
the significant interaction (i.e., scope of rivals’ im-
itation 3 product technology heterogeneity) in
Figures 3. Specifically, we show in Figure 3 the
actual scope of rivals’ imitation and product tech-
nology heterogeneity associated with various levels
of performance. As expected, the relationship be-
tween scope of rivals’ imitation and focal firm per-
formance is more negative for high levels of product
technology heterogeneity.

It is worth noting that although not directly
predicted in our theory, our regression analysis
offers interesting results on the impact of product
technology heterogeneity on firm performance and
on rivals’ imitative actions. Model 3 (Table 2) and
Figures 2 and 3 show that product technology
heterogeneity has a positive effect on focal firm

FIGURE 2
Scope of Focal Firm’s Imitation, Product Technology Heterogeneity, and Focal Firm Performance

3000

2

500

2000

1500

1000

500

0
0 1 2 3 0

1
2
3
4
Product technology
heterogeneity

Focal firm’s
imitation scope

F
o
ca

l
fi

rm
p

er
fo

rm
a
n

ce

2500–3000

2000–

2500

1500–2000

1000–1500

500–1000

0–500

2017 1903Giachetti, Lampel, and Li Pira

performance. In fact, when product technology
heterogeneity is high, products introduced by in-
dustry members are highly heterogeneous, and
direct competition is likely to be relatively weak,
since each firm in the industry attempts to carve
out its own unique product niche. Thus, although
we found that product technology heterogeneity
may create uncertainty and hamper the effective-
ness of focal firms’ imitative actions, overall firms
tend to achieve higher performance in this sce-
nario. As for the direct effect of product technology
heterogeneity on rivals’ imitative actions, we
found that the effect is positive on rivals’ imitation
scope (Model 6, Table 3), while the effect is nega-
tive on rivals’ average speed of imitation (Model 9,
Table 3): higher heterogeneity in product designs
triggers imitative responses aimed at catching the
opportunities offered by the variety of available
technologies, but the propensity to imitate several
different technologies limits the rivals’ ability to
imitate them rapidly.

Finally, although our paper looks at Red Queen
competition primarily from the point of view of
key competitive moves that involve imitation of
new product technologies, which in turn triggers
rivals’ imitative response (Figure 1), we also want
to take into account the possibility that rivals
respond to the focal firm’s imitative actions
with their own innovations. In Table 6, Models
14–16, we report the analysis of the effect of
a focal firm’s imitative actions on rivals’ innova-
tion scope. Given the excess of zero counts in
the rivals’ innovation scope dependent variable,

a zero-inflated Poisson regression was used
(Cameron & Trivedi, 2009). Model 16 is the full
model, also taking into account the moderating
effect of product technology heterogeneity. As can
be observed from Model 16, while the focal firm’s
average speed of imitation has no significant im-
pact on rivals’ innovation scope (b 5 0.05, p . .1),
the impact of the focal firm’s imitation scope is
negative and significant (b 5 20.20, p , .1). This re-
sult should be read together with the positive effect
of the focal firm’s imitation scope on rivals’ imita-
tion scope we found in Model 6: as the focal firm’s
imitative action (scope) increases, rivals tend to re-
spond with imitation at the expense of innovation.

Limitations and Avenues for Future Research

As may be expected, our study has limitations,
some of which create opportunities for future re-
search. First, as is the case in most empirical
studies in competitive dynamics, our study cap-
tures only observable strategies based on in-
formation reported in the press and industry trade
journals we examined. However, given the fact that
mobile phones regularly incorporate technologies
that originated in other product categories, such as
digital cameras, MP3 players, and video games, it is
likely that mobile phone vendors in our sample are
influenced by technological decisions made by
actors from other industries. This caveat applies to
the United Kingdom as well as global mobile
phone sales. Consequently, future research could

FIGURE 3
Rivals’ Scope of Imitation, Product Technology Heterogeneity, and Focal Firm Performance

3000
2500
2000
1500
1000
500

0
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 180 1

2 3
4

2500–3000

2000–2500

1500–2000
1000–1500
500–1000
0–500

Rivals’
imitation scope

Product technology
heterogeneity
F
o
ca
l
fi
rm
p
er
fo
rm
a
n
ce

1904 OctoberAcademy of Management Journal

examine how country and industry boundaries
influence Red Queen competitive imitation.

Second, our study examines an industry defined
by a single product, the mobile phone. Red Queen
competition in this case is likewise focused pri-
marily on improvements to this device. Empiri-
cally, studying an industry that is defined by
a single product is an advantage inasmuch as it
provides a context that allows us to examine Red
Queen competition with greater precision. How-
ever, this advantage is also a limitation, given the

fact that competition in many industries, for ex-
ample food retailing, is multi-product. It can be
reasonably expected that product line diversity will
produce different action–reaction dynamics than is
the case when competition is focused on a single
device. For example, the pressure to respond to
a rival’s move in one segment of the market may be
lower if the focal firm sees potential losses as minor
relative to the performance of its entire product
portfolio. Our results for imitation scope and speed
may be generalizable to other industries where

TABLE 6
Robust Zero-inflated Poisson Regression Analysis: Focal Firm Imitative Actions on Rivals’ Innovation Scope

Model 14 Model 15 Model 16

Rivals’ innovation
scope (t 1 1)
Rivals’ innovation
scope (t 1 1)
Rivals’ innovation
scope (t 1 1)

Constant 20.13 20.75** 20.85**
(–1.10) (–6.99) (–7.47)

Independent variables
Focal firm’s imitation scope (t) 20.05 20.201

(–0.55) (–1.69)
Focal firm’s average speed of imitation (t) 20.03 0.05

(–0.44) (0.62)
Rivals’ imitation scope (t 1 1) 0.51** 0.52**

(12.71) (12.57)
Rivals’ average speed of imitation (t 1 1) 0.22** 0.20*

(2.58) (2.36)
Product technology heterogeneity (t) 20.30** 20.29**

(–2.87) (–2.78)
Interactions
Focal firm’s imitation scope 3 Product technology
heterogeneity

0.19 1

(1.72)
Focal firm’s average speed of imitation 3 Product
technology heterogeneity

20.04
(–0.43)

Controls
Focal firm’s innovation scope (t) 20.01 20.03 20.03

(–0.20) (–0.61) (–0.66)
Relative market position (t) 0.03 0.02 0.01

(0.51) (0.38) (0.24)
Industry concentration (t) 0.11 0.07 0.09

(1.31) (0.68) (0.96)
GDP volatility (t) 0.081 0.04 0.05

(1.73) (0.56) (0.82)
2nd quarter year t (largest new product launch) 20.21 20.08 20.07

(–1.36) (–0.58) (–0.47)
3rd quarter year t (largest new product launch) 20.16 20.08 20.05

(–0.98) (–0.51) (–0.32)
4th quarter year t (largest new product launch) 20.18 20.06 20.03

(–1.15) (–0.38) (–0.18)
n 566 566 566
Log-likelihood 2568.95 2518.29 2515.55
Likelihood Ratio x2 7.49 180.11 187.05

Notes: Estimates are based on standardized variables; z-statistics in parentheses.
1p , 0.10
*p , 0.05

**p , 0.01

2017 1905Giachetti, Lampel, and Li Pira

single products drive competition, but may not be
generalizable for multi-product industries. Further
research is clearly needed to extend the findings of
our study to industries where competition engages
firms that offer consumers a wide range of products.

Third, although wecontendthat product technology
heterogeneity can affect the way firms learn from
the technology adoption decisions of rivals, and un-
dertake actions accordingly, scholars of organiza-
tional learning have identified a variety of learning
mechanisms—e.g., mimetic, vicarious, and experien-
tial (Baum et al., 2000; Haunschild & Miner, 1997;
Lieberman & Asaba, 2006)—that are not captured in
our theory and empirical analysis. Whether firms se-
lect one mode of learning over another depends on
their resource endowment and the time they can wait
before committing to a decision, with inevitable dif-
ferent impacts on the type and effectiveness of their
imitative actions. It would be useful for future research
to develop appropriate measures of different learning
modes,aswellastoprovideatheoreticalbasisforthese
measures.

Finally, analysis of Red Queen competition is
usually studied through the lens of inter-firm ri-
valry, with an interest in how firms react to each
other’s moves (Delacour & Liarte, 2012). However,
Derfus et al. (2008) suggested that it is also impor-
tant to see Red Queen competition as a link be-
tween micro and macro industry dynamics. They
noted, for instance, that new product introduction
moves may represent a “positive sum” game in
which the race to introduce products with more
features and better technologies can increase con-
sumer demand for the industry as a whole. Para-
doxically, therefore, Red Queen competition can
lead to a competitive stalemate at the level of in-
dividual firms, while at the same time producing
greater benefits for all to share. The same can be
said for technological change. Firms introduce
new products and new technologies in order to
retain their position, but in the process of doing so
they move the industry’s technological frontier
forward. In principle, we can therefore say that Red
Queen competition often plays an important role
in linking competitive interactions at the micro
industry level with macro industry dynamics
(Felin, Foss, & Ployhart, 2015). This linking role is
potentially a fruitful area of Red Queen competi-
tion research. Future research should therefore
examine how different types of Red Queen com-
petition impact the evolution of industries, and,
vice versa, how the evolution of industries shapes
Red Queen competition.

REFERENCES

Abrahamson, E., & Rosenkopf, L. 1993. Institutional and
competitive bandwagons: Using mathematical mod-
eling as a tool to explore innovation diffusion.
Academy of Management Review, 18: 487–517.

Agarwal, R., Sarkar, M. B., & Echambadi, R. 2002. The
conditioning effect of time on firm survival: An in-
dustry life cycle approach. Academy of Management
Journal, 45: 971–994.

Alpert, F. H., & Kamins, M. A. 1995. An empirical in-
vestigation of consumer memory, attitude, and per-
ceptions toward pioneer and follower brands. Journal
of Marketing, 59: 34–45.

Argyres, N., Bigelow, L., & Nickerson, J. A. 2015. Domi-
nant designs, innovation shocks, and the follower’s
dilemma. Strategic Management Journal, 36:
216–234.

Barkema, H. G., Baum, J. A., & Mannix, E. A. 2002. Man-
agement challenges in a new time. Academy of
Management Journal, 45: 916–930.

Barnett, W. P., & Hansen, M. T. 1996. The red queen in
organizational evolution. Strategic Management
Journal, 17(S1): 139–157.

Barnett, W. P., & McKendrick, D. G. 2004. Why are some
organizations more competitive than others? Evi-
dence from a changing global market. Administrative
Science Quarterly, 49: 535–571.

Barnett, W. P., & Sorenson, O. 2002. The Red Queen in
organizational creation and development. Industrial
and Corporate Change, 11: 289–325.

Baucells, M., Weber, M., & Welfens, F. 2011. Reference
point formation and updating. Management Science,
57: 506–519.

Baum, J. A. C., & Ingram, P. 1998. Survival-enhancing
learning in the Manhattan hotel industry. Manage-
ment Science, 44: 996–1016.

Baum, J. A. C., Li, S. X., & Usher, J. M. 2000. Making the next
move: How experiential and vicarious learning shape
the locations of chains’ acquisitions. Administrative
Science Quarterly, 45: 766–801.

Baumol, W. J. 2004. Red-Queen games: Arms races, rule of
law and market economies. Journal of Evolutionary
Economics, 14: 237–247.

Bayus, B. L., & Agarwal, R. 2007. The role of pre-entry
experience, entry timing, and product technology
strategies in explaining firm survival. Management
Science, 53: 1887–1902.

Cameron, C. A., & Trivedi, P. K. 2009. Microeconometrics
using Stata. College Station, TX: Stata Press.

Cameron, T. A. 2005. Updating subjective risks in the
presence of conflicting information: an application to

1906 OctoberAcademy of Management Journal

climate change. Journal of Risk and Uncertainty, 30:
63–97.

Carroll, L. 1960. Through the looking glass and what
Alice found there: The annotated Alice. New York,
NY: Bramhall House.

Carpenter, G. S., & Nakamoto, K. 1989. Consumer prefer-
ence formation and pioneering advantage. JMR,
Journal of Marketing Research, 26: 285–298.

Chatterjee, S., & Hadi, A. S. 2006. Regression analysis by
example. New York, NY: Wiley.

Chen, M. J., Lin, H. C., & Michel, J. G. 2010. Navigating in
a hypercompetitive environment: the roles of action
aggressiveness and TMT integration. Strategic Man-
agement Journal, 31: 1410–1430.

Chen, M. J., & Miller, D. 1994. Competitive attack, retaliation
and performance: an expectancy‐valence framework.
Strategic Management Journal, 15: 85–102.

Chen, M. J., & Miller, D. 2012. Competitive dynamics:
Themes, trends, and a prospective research plat-
form. The Academy of Management Annals, 6:
135–210.

Chen, Y., & Turut, O. 2013. Context-dependent preferences
and innovation strategy. Management Science, 59:
2747–2765.

Christensen, C. M., & Bower, J. L. 1996. Customer power,
strategic investment, and the failure of leading firms.
Strategic Management Journal, 17: 197–218.

Cohen, W. M., & Levinthal, D. A. 1989. Innovation and
learning: The two faces of R&D. Economic Journal
(Oxford), 99: 569–596.

Csaszar, F. A., & Siggelkow, N. 2010. How much to copy?
Determinants of effective imitation breadth. Organi-
zation Science, 21: 661–676.

D’Aveni, R. A. 1994. Hypercompetition: Managing the
dynamics of strategic maneuvering. New York, NY:
The Free Press.

Delacour, H., & Liarte, S. 2012. The Red Queen Effect:
Principles, synthesis and implications for strategy.
M@n@gement, 15: 313–330.

Derfus, P. J., Maggitti, P. G., Grimm, C. M., & Smith, K. G.
2008. The Red Queen effect: Competitive actions and
firm performance. Academy of Management Jour-
nal, 51: 61–80.

Ethiraj, S. K., & Zhu, D. H. 2008. Performance effects of
imitative entry. Strategic Management Journal, 29:
797–817.

Euromonitor International (2003–2008). Mobile phones in
the U.K. Annual directories. London, U.K.: Euro-
monitor International Ltd.

Felin, T., Foss, N. J., & Ployhart, R. E. 2015. The micro-
foundations movement in strategy and organization

theory. The Academy of Management Annals, 9:
575–632.

Ferrier, W. J., Smith, K. G., & Grimm, C. M. 1999. The role of
competitive action in market share erosion and in-
dustry dethronement: A study of industry leaders and
challengers. Academy of Management Journal, 42:
372–388.

Frenken, K., Saviotti, P. P., & Trommetter, M. 1999. Va-
riety and niche creation in aircraft, helicopters, mo-
torcycles and microcomputers. Research Policy, 28:
469–488.

Gaba, V., & Terlaak,A. 2013. Decomposing uncertainty and
its effects on imitation in firm exit decisions. Orga-
nization Science, 24: 1847–1869.

Giachetti, C., & Dagnino, G. B. 2017. The impact of tech-
nological convergence on firms’ product portfolio
strategy: An information-based imitation approach.
R&D Management, 47: 17–35.

Giachetti, C., & Lanzolla, G. 2016. Product technology
imitation over the product diffusion cycle: Which
companies and product innovations do competitors
imitate more quickly? Long Range Planning, 46:
250–264.

Giachetti, C., & Marchi, G. 2017. Successive changes in
leadership in the worldwide mobile phone in-
dustry: The role of windows of opportunity and
firms’ competitive action. Research Policy, 46:
352–364.

Greve, H. R. 1996. Patterns of competition: The diffusion of
a market position in radio broadcasting. Administrative
Science Quarterly, 41: 29–60.

Greve, H. R. 1998. Managerial cognition and the mimetic
adoption of market positions: What you see is what
you do. Strategic Management Journal, 19:
967–988.

Haddow, A., Hare, C., Hooley, J., & Shakir, T. 2013. Mac-
roeconomic uncertainty: What is it, how can we
measure it and why does it matter? Bank of England
quarterly bulletin, June, Q2.

Haunschild, P. R., & Miner, A. S. 1997. Modes of in-
terorganizational imitation: The effects of outcome
salience and uncertainty. Administrative Science
Quarterly, 42: 472–500.

Honoré, B. E. 1992. Trimmed LAD and least squares es-
timation of truncated and censored regression
models with fixed effects. Econometrica, 60:
533–565.

Hsieh, K. Y., & Vermeulen, F. 2013. The structure of com-
petition: How competition between one’s rivals in-
fluences imitative market entry. Organization
Science, 25: 299–319.

Kardes, F. R., & Kalyanaram, G. 1992. Order-of-entry
effects on consumer memory and judgment: An

2017 1907Giachetti, Lampel, and Li Pira

information integration perspective. Journal of
Marketing Research, 29: 343–357.

Krishnan, V., & Bhattacharya, S. 2002. Technology selec-
tion and commitment in new product development:
The role of uncertainty and design flexibility. Man-
agement Science, 48: 313–327.

Lampel, J., & Shamsie, J. 2009. All that running: Red Queen
competition inthe U.S. motion picture industry. Paper
presented at the Academy of Management conference,
Chicago, August 2009.

Lee, H., Smith, K., Grimm, C. M., & Schomburg, A. 2000.
Timing, order and durability of new product advan-
tages with imitation. Strategic Management Journal,
21: 23–30.

Levitt, T. 1966. Innovative imitation. Harvard Business
Review, 44: 63–70.

LiCalzi, M., & Marchiori, D. 2013. Pack light on the move:
Exploitation and exploration in a dynamic environ-
ment. In: S. Leitner & F. Wall (eds.), Artificial eco-
nomics and self organization: 205–216. Berlin:
Springer.

Lieberman, M. B., & Asaba, S. 2006. Why do firms imitate
each other? Academy of Management Review, 31:
366–385.

Lieberman, M. B., & Montgomery, D. B. 1988. First-mover
advantages. Strategic Management Journal, 9(S1):
41–58.

Lieberman, M. B., & Montgomery, D. B. 1998. First-mover
(dis)advantages: Retrospective and link with the
resource-based view. Strategic Management Jour-
nal, 19: 1111–1125.

Lippman, S. A., &Rumelt, R. P. 1982. Uncertain imitability:
An analysis of interfirm differences in efficiency un-
der competition. The Bell Journal of Economics, 13:
418–438.

Madhok, A., Li, S., & Priem, R. L. 2010. The resource‐based
view revisited: Comparative firm advantage, willingness‐
based isolating mechanisms and competitive het-
erogeneity. European Management Review, 7:
91–100.

Makadok, R. 1998. Can first-mover and early-mover ad-
vantages be sustained in an industry with low barriers
to entry/imitation? Strategic Management Journal,
19: 683–696.

Markides, C. C., & Geroski, P. A. 2004. Fast second: How
smart companies bypass radical innovation to enter
and dominate new markets. San Francisco, CA: John
Wiley & Sons.

Miller, D., & Chen, M. J. 1994. Sources and consequences of
competitive inertia: A study of the US airline industry.
Administrative Science Quarterly, 39: 1–23.

Mintel International Group Limited (1997–2008). Mobile
phones and network providers in the U.K. Annual
Directories. London, U.K.: Mintel International
Group Limited.

Murmann, J. P., & Frenken, K. 2006. Toward a systematic
framework for research on dominant designs, tech-
nological innovations, and industrial change. Re-
search Policy, 35: 925–952.

Narasimhan, C., & Turut, O. 2013. Differentiate or imitate?
The role of context-dependent preferences. Market-
ing Science, 32: 393–410.

Ndofor, H. A., Sirmon, D. G., & He, X. 2011. Firm resources,
competitive actions and performance: investigating
a mediated model with evidence from the in-vitro di-
agnostics industry. Strategic Management Journal,
32: 640–657.

Nelson, R., & Winter, S. 1982. An evolutionary theory of
economic change. Cambridge, MA: Harvard Univer-
sity Press.

O’Shaughnessy, J. 1989. Why people buy. Oxford, U.K.:
Oxford University Press.

Pessemier, E. A. 1978. Stochastic properties of changing
preferences. The American Economic Review, 68:
380–385.

Posen, H. E., Lee, J., & Yi, S. 2013. The power of imperfect
imitation. Strategic Management Journal, 34:
149–164.

Posen, H. E., & Levinthal, D. A. 2012. Chasing a moving
target: Exploitation and exploration in dy-
namic environments. Management Science, 58:
587–601.

Rao, R., & Rutenberg, D. 1979. Pre-empting an alert rival:
strategic timing of the first plant by analysis of so-
phisticated rivalry. The Bell Journal of Economics,
10: 412–428.

Rhee, M., Kim, Y. C., & Han, J. 2006. Confidence in imita-
tion: Niche-width strategy in the U.K. automobile in-
dustry. Management Science, 52: 501–513.

Rindova, V., Ferrier, W. J., & Wiltbank, R. 2010. Value from
gestalt: how sequences of competitive actions create
advantage for firms in nascent markets. Strategic
Management Journal, 31: 1474–1497.

Ross, J. M., & Sharapov, D. 2015. When the leader follows:
Avoiding dethronement through imitation. Academy
of Management Journal, 58: 658–679.

Schumpeter, J. 1942. Capitalism, socialism, and de-
mocracy. New York, NY: Harper & Row.

Semadeni, M., & Anderson, B. S. 2010. The follower’s di-
lemma: Innovation and imitation in the professional
services industry. Academy of Management Journal,
53: 1175–1193.

1908 OctoberAcademy of Management Journal

Shannon, C. E. 1948. A mathematical theory of communi-
cation. Bell System Technical Journal, 27, 623–656.

Smith, K. G., Ferrier, W. J., & Ndofor, H. 2001. Competitive
dynamicsresearch:Critiqueandfuturedirections.In:M.
Hitt, R.E. Freeman, & J. Harrison (Eds.), Handbook of
Strategic Management, 315–361. London: Blackwell.

Smith, K. G., Grimm, C., Gannon, M., & Chen, M. J. 1991.
Organizational information processing, competitive re-
sponses and performance in the U.S. domestic airline
industry. Academy of Management Journal, 34: 60–85.

Teece,D.J.1998.Capturingvaluefromknowledgeassets:The
new economy, markets for know-how, and intangible
assets. California Management Review, 40: 55–79.

Utterback, J., & Suarez, F. F. 1993. Innovation, competition
and industry structure. Research Policy, 22: 1–21.

Van Valen, L. 1973. A new evolutionary law. Evolutionary
Theory, 1: 1–30.

Waldman, D. E., & Jensen, E. J. 2012. Industrial organi-
zation: Theory and practice. Upper Saddle River,
N.J.: Prentice Hall.

Wiggins, R. R., & Ruefli, T. W. 2005. Schumpeter’s ghost: Is
hypercompetition making the best of times shorter?
Strategic Management Journal, 26: 887–911.

Young, G., Smith, K. G., & Grimm, C. M. 1996. “Austrian”
and industrial organization perspectives on firm-level
competitive activity and performance. Organization
Science, 7: 243–254.

Zellner, A. 1962. An efficient method of estimating seem-
ingly unrelated regressions and tests for aggregation

bias. Journal of the American Statistical Associa-

tion, 57: 348–368.

Zhou, J. 2011. Reference dependence and market compe-

tition. Journal of Economics & Management Strat-

egy, 20: 1073–1097.

Claudio Giachetti (claudio.giachetti@unive.it) is an associate
professor of strategy at Ca’ Foscari University of Venice (Italy),
Department of Management. He received his PhD from Ca’
Foscari University of Venice. His primary research interests
concern competitive dynamics and product innovation in
rapidlychangingtechnologicalandinstitutionalenvironments.

Joseph Lampel (joseph.lampel@manchester.ac.uk) is
Eddie Davies Professor of Enterprise and Innovation
Management at Alliance Manchester Business School,
University of Manchester (U.K.). He received his PhD
from McGill University, Montreal. His research fo-
cuses on the dynamics of competition, innovation de-
cision making, and strategy formation in creative
industries.

Stefano Li Pira (stefano.li-pira@wbs.ac.uk) is an assistant
professor at Warwick Business School, the University of
Warwick (U.K.). He received his PhD from Ca’ Foscari
University of Venice. His primary research interests con-
cern competitive dynamics and imitation in technology
intensive industries.

2017 1909Giachetti, Lampel, and Li Pira

mailto:claudio.giachetti@unive.it

mailto:joseph.lampel@manchester.ac.uk

mailto:stefano.li-pira@wbs.ac.uk

APPENDIX A

FIGURE A1
Average Number of Innovations and Imitations by U.K. Mobile Phone Vendors (1997–2007)

Total number of technologies

42
40
38
36
34
32
30
28
26
24
22
20
18
16
14
12
10
8
6
4
2
0

7
6
5
4
3
2
1

0
1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007

Imitations

F
ir

m
’s

a
v

er
a

ge
n

u
m

b
er

o
f

im
it
a
ti
o
n

s
a

n
d

i
n

n
o

v
a

io
n
s

Innovations Technologies per product

A
v

er
a
ge
n
u
m
b
er
o
f
te
ch
n
o

lo
gi

es
p

er
p

ro
d
u
ct

;
T

o
ta

l
n

u
m
b
er
o
f
te
ch
n
o
lo
gi

es
c

u
rr

en
tl

y
a

d
o
p

te
d

Notes: Values presented in the figure are based only on “regular phones”—smartphones are excluded. The average number of innovations
(imitations) expresses, on average, in a given year, how many new product technologies are introduced (imitated) by handset vendors. The
average number of technologies per product refers to the average number of technologies that handset vendors installed in their phones in
a given year. The total number of technologies refers to the total number of different technologies that were adopted in a given year by handset
vendors.

1910 OctoberAcademy of Management Journal

TABLE A1
Product Technologies Introduced in the U.K. Mobile Phone Industry from 1997–2007

Technological
system Types of functions offered

List of technologies (month of
introduction in the U.K. market) Description

Networking Mobile phone networks use signals
on specific frequency bands.
Phones must be compatible with
these bands in order to work with
the network.

Dual band (Feb 1998) Phone’s ability to work with two of
thefourmajor GSM(GlobalSystem
for Mobile Communications)
frequency bands. An important
feature for users who wish to use
the same handset in different
locations where the networks work
on different bands. For example,
some European dual-band phones
do not work in the U.S., and vice
versa.

Tri band (Aug 1999) Phone’s ability to work with three of
the four major GSM frequency
bands, allowing it to work in most
parts of the world.

Quad band (Oct 2003) Phone’s ability to work with the four
major GSM frequency bands (850/
900/1800/1900 MHz), making it
compatible with all the major GSM
networks in the world.

Wideband Code Division Multiple
Access (WCDMA) (Mar 2003)

Third-generation (3G) wireless
standard that allows use of both
voice and data. It has different
frequency bands (Europe and
Asia—2100MHz, North
America—1900MHz and
850MHz).

High-speed data
transfer

Mobile phone networks support
different types of data transfer,
which allows users to access
mobile internet, MMS and other
advanced features like video
streaming.

High-Speed Circuit-Switched Data
(HSCSD) (Nov 2000)

System for data calls on GSM
networks that came before packet-
based systems such as GPRS and
EDGE. It was never widely adopted
outside Europe.

General Packet Radio Service (GPRS)
(Mar 2001)

A packet-switching technology that
enables data transfers through
cellular networks. It is used for
mobile internet, MMS and other
data communications. Informally,
GPRS is also called 2.5G.

Enhanced Data rates for GSM
Evolution (EDGE) (Feb 2004)

Data system used on top of GSM
networks. It provides nearly three
times faster speeds than the
outdated GPRS system. EDGE
meets the requirements for a 3G
network but is usually classified as
2.75G.

Universal Mobile
Telecommunications System
(UMTS) (Mar 2003)

Includes high data speeds (2 Mbps),
always-on data access, and greater
voice capacity, enabling such
advanced features as live video
streaming.

High Speed Downlink Packet Access
(HSDPA) (Mar 2007)

Upgrade for UMTS networks that
doubles network capacity and
increases download data speeds by
five times or more.

Phone call Phone call functionalities refer to the
way the user can make a phone call
(e.g., voicedialing the number),the

Vibrate alert (Jan 1997) Can alert user to events such as an
incoming call or an incoming
message with a vibrate alert.

2017 1911Giachetti, Lampel, and Li Pira

TABLE A1
(Continued)

Technological
system Types of functions offered
List of technologies (month of
introduction in the U.K. market) Description

type of call (i.e., voice vs. video),
and the type of call alert (the
mobile phone can alert the user to
events such as an incoming call or
an incoming message in a number
of ways).

Voice Dial (Jul 1997) Allows the user to dial a number via
a voice command.

Polyphonic ringtones(Jan 2000) Creates realistic-sounding music by
synthesizing several notes
simultaneously. The more notes
the synthesizer can play
simultaneously, the richer the
musical effect. Usually mobile
phone synthesizers can reproduce
from 4 to 72 simultaneous tones.

True tones (Feb 2003) Audio recordings, typically in
a common format such as MP3,
AAC, or WMA.

Downloadable ringtones (Feb 1998) Allows the user to load a new
ringtone by downloading via
a special SMS/MMS, or from the
Internet.

Composer (Aug 1997) Allows the user to create musical
notes and then produce
a customized ringtone.

Recordable (Jan 2000) Permits sound recording—e.g., of
someone’s voice—and then using
it as a ringtone.

Video Call (Mar 2003) 3G-network feature that allows two
callers to talk to each other while at
the same time viewing live video
form each other’s phone.

Connectivity Protocols for exchanging data over
short distances from fixed and
mobile devices, creating personal
area networks.

Infrared (Oct 1997) Standard for transmitting data using
an infrared port. Uses a beam of
infrared light to transmit
information and so requires direct
line of sight and operates only at
close range.

Bluetooth (Aug 2001) Wireless protocol for exchanging
data over short distances from
fixed and mobile devices, creating
personal area networks.

Universal Serial Bus (USB) (Sept
2001)

Standard for a wired connection
between two electronic devices,
including a mobile phone and
a desktop computer. The
connection is made using a cable
that has a connector at either end.

Messaging In addition to pure voice calls,
messaging has been a core service
since the beginning of GSM mobile
telephony.

Enterprise Messaging System (EMS)
(Aug 1999)

Extension of SMS (Short Message
Service), which allows mobile
phones to send and receive
messages that have special text
formatting, animations, graphics,
sound effects, and ringtones. It is
an intermediate technology
between SMS and rich multimedia
messages (MMS).

Multimedia Messaging Service
(MMS) (May 2002)

Store-and-forward messaging service
that allows subscribers to
exchange multimedia files as
messages (text, picture, audio,

1912 OctoberAcademy of Management Journal

TABLE A1
(Continued)
Technological
system Types of functions offered
List of technologies (month of
introduction in the U.K. market) Description

video, or a combination). In order
to send or receiveanMMS,the user
must have a compatible phone that
is running over a GPRS or 3G
network.

SMS chat (Nov 2000) Analogous to the pervasive use of
SMS as a type of instant messaging,
much like chatting on a computer.
The threaded message or
conversation-style layout displays
the incoming and outgoing
messages between two
participants in a single pane
ordered chronologically.

Instant Messaging (IM) (May 2002) Ability to engage in instant
messaging services from a mobile
handset. Mobile IM allows users to
address messages to others using
a dynamic address book full of
users, with their online status
updated constantly. Permits
anyone participating to know
when their “buddies” are available
for chat. Mobile IM is viewed as
a logical extension of the popular
SMS service.

E-mail (Mar 1998) Some phones provide a full e-mail
client that can connect to a public
or private e-mail server. There are
different protocols used by the
servers and some may not be
supported by the phone’s e-mail
client.

Display Display is one of the most relevant
aesthetic features of the mobile
phone. Size, color, and physical
interaction have a strong influence
on the user’s experience.

Colorscreen: 4 colors (Sep 1997), 256
colors (Dec 2001), 4 K color (Jun
2002), 65 K colors (Nov 2002),
256 K colors (May 2004), 16 MK
color (Aug 2005)

Display is able to produce a number
of different colors. A higher
number results in a broader range
of distinct colors. We identified six
levels of color screen.

Display shape: Display Vertical (May
1998), Display Squared (Nov 2000)

Mobile phone display shape that is
convenient for the different
function supported (messaging,
photos, etc.). We identified two
categories based on the display
width/height ratio (squared
display, vertical display).

Touchscreen (May 1998) Display that responds to direct touch
manipulation, either by finger,
stylus, or both.

Technological
convergences

Technologies traditionally
originating in other industries, and
“converging” into the mobile
phone industry.

Photo camera (Aug 2002)
Videocamera (Mar 2003)

Camera that can function as a digital
camera, and in some cases can also
shoot video.

Photo resolution: 1 Mp 2 Mp (Oct
2004), 2 Mp 3 Mp (Jun 2005), 3 Mp
4 Mp (Sep 2006)

Indicates the number of pixels on
a display or in a camera sensor
(specifically in a digital image). A
higher resolution means more
pixels and more pixels provide the

2017 1913Giachetti, Lampel, and Li Pira

TABLE A1
(Continued)
Technological
system Types of functions offered
List of technologies (month of
introduction in the U.K. market) Description

ability to display more visual
information (resulting in greater
clarity and more detail).

Voice memo (Jan 1997) Permitsuserstorecordanotethatcan
be heard whenever and wherever
necessary. Some devices limit the
duration of such memos, whereas
others allow recording until they
run out of memory.

MP3 (Dec 2000) Audio storage protocol that stores
music in a compressed format with
very little loss in sound quality.
MP3 files can be played using the
music player of the mobile phone
or set as a ringtone.

Internet capabilities: HDML (Mar
1998); WML (Aug 1999); HTML
(Mar 1998); XHTML (Nov 2002)

Various markup languages(ML) have
been introduced to allow the
handsetto surf theInternet.Mostof
them allow only access to
simplified Internet pages.

Document viewer (Jul 2005) Program for displaying MS Word,
Excel, and PowerPoint files.

FM Radio (Apr 2000) Permits user to listen to most live-
broadcast FM radio stations.
Almost all phones with an FM
radio tuner require a wired headset
to be connected to the unit as it is
used as an antenna.

Games (Jan 1998) Many phones include simple games
for the user to pass the time. The
games referred to here are
preinstalled on the phone and do
not require a wireless connection
to play.

Notes: Definitions and technical descriptions of the sampled technologies were collected from both the special-interest magazines used for
the analysis, and online catalogs such as www.gsmarena.com. Information on the month of introduction of a new technology in the U.K. market
was collected from the special-interest magazines used for our analysis.

1914 OctoberAcademy of Management Journal

http://www.gsmarena.com

Copyright of Academy of Management Journal is the property of Academy of Management
and its content may not be copied or emailed to multiple sites or posted to a listserv without
the copyright holder’s express written permission. However, users may print, download, or
email articles for individual use.