Business Strategy and Forecasting as Competitive Advantages

r Academy of Management Journal 2017, Vol. 60, No. 5, 1882–1914. https://doi.org/10.5465/amj.2015.0295

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

CLAUDIO GIACHETTI Ca’ Foscari University of Venice

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 (as 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.

1882

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.

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

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 ta 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 io 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

In 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

2500

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 ([email protected]) 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 ([email protected]) 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 ([email protected]) 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

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

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.