THE FOLLOWER’S DILEMMA: INNOVATION AND · PDF fileimitator differentially weighs the...

THE FOLLOWER’S DILEMMA: INNOVATION AND IMITATIONIN THE PROFESSIONAL SERVICES INDUSTRY

MATTHEW SEMADENIBRIAN S. ANDERSON

Indiana University

Firm decision makers contemplating imitation of a competitor’s innovation face adilemma: Imitate a new, unproven offering, or forgo imitation and perhaps miss out onthe “next big thing”? Approaching this underexplored area of research, we applyinformation-based imitation theory to evaluate organization- and offering-level char-acteristics influencing imitation under conditions of high environmental uncertaintyand high information asymmetry. In analyzing the service mark filings of the 50 largestmanagement consulting firms over 11 years, we find that although an innovator’sorganization-level characteristics increase imitation, offering-level characteristics de-crease imitation. Furthermore, organization- and offering-level characteristics inter-act, resulting in different imitation outcomes.

Few topics in the strategy literature have beenexamined as thoroughly as the actions of firstmover firms (e.g., Lambkin, 1988; Lieberman &Montgomery, 1988; Makadok, 1998; Suarez & Lan-zolla, 2007). The argument of the first mover per-spective is that firms that proactively enter a newproduct or market space may enjoy a temporal com-petitive advantage over follower firms (Suarez &Lanzolla, 2007). Research has shown, however, thatfirst mover firms in fact may be as likely to sufferfrom competitive disadvantages as to obtain com-petitive advantages (Boulding & Christen, 2008;Christensen, 1997; Lieberman & Montgomery,1998). Therein lies the risk of going first: dependingon context, a firm’s innovation has an almost evenprobability of success or failure (Suarez & Lanzolla,2007). Competitors of the first mover, however, facea dilemma of their own. On the one hand, such afirm may choose to imitate all or a portion of thefirst mover’s innovation. As a competitive re-sponse, such mimicry represents an effort to mini-mize any advantage the first mover may have de-rived from the innovation (DiMaggio & Powell,1983; Makadok, 1998). On the other hand, the firmmay choose to bypass the innovation and not imi-tate, acting perhaps on a belief that the innovationis too risky or the environment too uncertain(Dougherty & Heller, 1994). Regardless of thechoice the follower firm makes—to imitate or not—both options contain considerable uncertainty.

There is the possibility, for example, that the firstmover made a wrong decision and the innovationwill not be well received by the market. Under thisscenario, the imitator would be simply repeatingthe mistake made by the first mover. Conversely,there is the possibility that the first mover has hitupon the “next big thing” and, by not imitating theinnovation, the competitor firm may miss a signif-icant market opportunity. There is also the possi-bility that imitation will spark an increased com-petitive rivalry, thereby lowering industry-wideprofit potential. Not imitating the first mover, how-ever, may threaten the competitor’s legitimacy; thatis, it may be perceived as staid and not in tune withchanges in the market. Thus, a “follower’s di-lemma” is inherent in imitation decisions; firmsmust weigh the uncertainty of imitating a firstmover’s offering against the uncertainty of notimitating.

Although prior research provides guidance onenvironmental and organizational antecedents toimitation (e.g., Abrahamson, 1996; Lieberman &Asaba, 2006; Terlaak & King, 2007), significant gapsremain in understanding the causes and conse-quences of imitation decisions. For example, theuse of imitation as a competitive response in envi-ronmental contexts where competitors possess sim-ilar market knowledge is well understood (e.g.,Chen, Su, & Tsai, 2007; Ferrier, 2001; Gimeno,1999), but less understood is imitation under con-ditions of environmental uncertainty and highinformation asymmetry (i.e., when market actorspossess different and unequal stores of marketknowledge). Lieberman and Asaba made an impor-tant observation about this environmental con-text—the confluence of environmental uncertainty

We thank Jeff Covin and Tieying Yu for their com-ments on drafts. We also thank David Ketchen and threeanonymous reviewers for the guidance that theyprovided.

� Academy of Management Journal2010, Vol. 53, No. 5, 1175–1193.

1175

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 expresswritten permission. Users may print, download or email articles for individual use only.

and high information asymmetry—stating that insuch a setting “imitation processes are most inter-esting [because] . . . few decisions have outcomesthat are fully predictable” (2006: 366). To furtheremphasize Lieberman and Asaba’s (2006) argu-ment, researching imitation in this environmentalcontext is both interesting and important because,though imitation is a prevalent competitive tactic,in highly uncertain environments in which infor-mation asymmetry is high, outcomes, favorable orunfavorable, are uncertain.

From an empirical perspective, another gap inthe imitation literature lies in the operation-alization of the imitation and innovation con-structs. Previous research has tended to portrayinnovation and imitation as dichotomous variables,framing innovation as either present or not, andfirms as either imitating or not imitating (e.g.,Greve, 1998; Guler, Guillen, & MacPherson, 2002;Haunschild, 1993; Lee, Lee, & Rho, 2002; Makadok,1998). What is lacking is a finer-grained analysis ofhow degree of innovation influences degree of im-itation, or, stated differently, the question ofwhether the level of imitation varies with the levelof innovation.

We seek to address the two knowledge gaps pos-ited above and make two additional contributionsto the imitation and innovation literatures. First,we argue that imitation is predicated on both or-ganization- and offering-level characteristics of afirst mover and its innovation, and the interactionof these elements. A multilevel view of the imita-tion decision is consistent with the notion that animitator differentially weighs the characteristics ofthe first mover and the characteristics of the inno-vation when considering whether or not to imitateand provides finer-grained insight into the anteced-ents of imitation.

Second, the setting for this research is the ser-vices sector, which comprises over half of theAmerican economy and is still growing, accordingto the U.S. Census Bureau. Prior research on imita-tion has largely focused on product or commoditymarkets. Bowen and Ford (2002) suggested that theintangibility of service-based markets creates anumber of competitive differences between prod-uct-focused firms and those primarily offering ser-vices. Given the growing importance of the servicessector, as well as the potential for conclusions dif-fering from those found in product-market settings,studies employing service sector data are inher-ently valuable to the strategy field.

It is also important to note the positioning of thisresearch relative to the competitive dynamics liter-ature. Like research on the first mover phenome-non, the study of competitive dynamics has an

extensive history in the strategic management field(e.g., Chen, 1996; Chen & MacMillan, 1992; Porter,1980), and it continues to be of interest to strategyscholars, particularly the work on multimarketcompetition (e.g., Chen et al., 2007; Gimeno, 1999).Generally speaking, the argument of multimarkettheory is that competitor firms may exhibit differ-ent competitive behaviors in each market in whichthey compete (Chen, 1996). Two assumptions un-derlying multimarket theory concern factors pos-ited to in part determine whether a competitiveaction pursued in one market can affect the com-petitive action pursued in another market.

The first is the assumption of competitive rivalry;that is, two firms’ leaderships consider each otheras competitors in multiple market contexts and areengaged in some manifestation of “challenge-re-sponse” behavior (Chen et al., 2007). To describethis assumption another way, for a competitive ac-tion in one market context to affect competitiveaction in another, the competitors must have atleast some degree of rivalry in both markets. Thesecond assumption is that each competitor mustunderstand the consequence of a particular com-petitive action, at least to some degree (Chen,1996). For example, in the airline industry, if air-line A lowers its price on a route important toairline B, decision makers at both airlines are cog-nizant of the consequences of such a competitiveaction.

Both of these assumptions are pertinent for thecurrent study, and in particular, for framing how itdiffers from research into multimarket competitivedynamics. The objective of this study was to ex-plore the antecedents to imitation of an innovationunder conditions of environmental uncertainty andhigh information asymmetry, where the conse-quences of innovation and imitation are largelyunknowable. In short, although multimarket com-petitive theory does address a number of competi-tive situations, it does not provide a clear theoret-ical lens through which to evaluate imitation in theenvironmental context of interest in the currentstudy—a context that is becoming far more preva-lent in industry (Lieberman & Asaba, 2006).

THEORETICAL FRAMEWORK

At the outset of presenting our theoretical frame-work, we wish to note that we use the term firm toidentify a focal firm (i.e., the firm that is consider-ing imitation), and the term competitor to identifythe organization whose services the focal firm isconsidering imitating (i.e., the first mover). Use ofthis naming convention continues throughout ourarticle.

1176 OctoberAcademy of Management Journal

Innovation and Imitation

Abrahamson (1996) suggested that, in introduc-ing any new product or service, firms must balancethe norms of rationality against the norms of pro-gressiveness. Rationality dictates that firms’ offer-ings appear reasonable and in-step with the market(i.e., not implausible or outrageous) to gain accep-tance. Conversely, per the norm of progressiveness,firms must be fresh and innovative, lest they appearstaid and lagging behind changing market condi-tions. Whereas highly uncertain environments tendto favor innovative, progressive behavior (Covin &Slevin, 1991; Miller, 1983), the opposing force ofrationality suggests that even in such environ-ments, innovative behavior should be tempered.Indeed, being too innovative can also threaten or-ganizational legitimacy (Deephouse, 1999) andmay, in fact, imperil firm performance (Levinthal &March, 1993; March, 1991). Thus, firms must bal-ance the need to be innovative with the need toappear reasonable and credible in their offer-ings—in other words, to be new, but not too new.Finding balance between these norms is at the heartof the tension between innovation and imitation:innovation generally advances progressiveness, butimitation reinforces rationality.

The need for balance between these two normssuggests firms that tend to favor innovation mayalso pursue imitation in some circumstances and,conversely, firms that favor imitation may also en-gage in selective innovation. Stated differently,firms are not unidimensional in their approaches toinnovation or to imitation, and they find balancebetween the norms by pursuing both behaviors(Chen, 1996). Furthermore, there is also the possi-bility of seeking balance between the norms withinan imitation decision itself. For example, imitatorsmay seek to advance their progressiveness throughimitating all, or almost all, of an innovation,thereby attempting to “ride the coattails” of the firstmover (Abrahamson, 1996). Conversely, prospec-tive imitators may chose to imitate only a smallportion of an innovation, attempting to advancerationality by keeping pace with market changeswhile also attempting to limit downside risk froman untested innovation. The range of response op-tions available to potential imitators raises thisquestion: Why, and under what conditions, will afirm choose imitation, and if it does, to what degreewill it imitate a competitor’s innovation?

Theories of Information and Rivalry

In their recent survey of the imitation literature,Lieberman and Asaba (2006) put forth two theoret-

ical frameworks for imitation decision making—one based in information and the other in rivalry.

Information-based imitation. The information-based imitation framework arises from the fields ofeconomics and sociology and casts uncertainty as anecessary condition. The basic tenet of economicinformation-based imitation is the idea that in un-certain environments, market knowledge is hetero-geneous (information asymmetry is high), and thecompetitor introducing an innovation may be per-ceived to have superior market knowledge (Bikh-chandani, Hirshleifer, & Welch, 1992). Further-more, Abrahamson (1996) and Bikhchandani et al.(1992) suggested that firms are more likely to fol-low competitors that are perceived to be “fashionleaders” or “trend setters”—a status that impartssignificance to their decisions. The key point of thefashion leaders concept (see Abrahamson, 1991,1996) is that a competitor with a reputation forbeing at the forefront of innovation, or with a par-ticular competence in a given area, is often as-sumed to possess superior market knowledge(Lieberman & Asaba, 2006). Thus, under informa-tion-based imitation, firms will actively look forand evaluate the signals sent by their competitorsfor information concerning the competitors’ beliefsabout the market; such signals are typically con-veyed by the competitors’ historical reputation anddemonstrated competencies (Bikhchandani, Hir-shleifer, & Welch, 1998).

Nonetheless, evaluating the quality of this signalis difficult, a fact that often causes firms to givegreater weight to signals that are supported by spe-cific competitor commitments (Bikhchandani et al.,1998). For example, competitor commitments thatare accompanied by a significant investment on itspart, and/or also strategic changes, increase thecredibility of the signals and thereby their salience(Lee, 2001; Milgrom & Roberts, 1986; Williamson,1983). Hence, when a competitor makes a substan-tive commitment—such as introducing a new prod-uct or service—this commitment conveys specific,credible information to the potential imitator aboutthe competitor’s view of market conditions. In pars-ing the signal, the potential imitator will assess notonly what it knows about the signal sender, butalso the information contained in the new prod-uct or service introduction, in line with the no-tion of multilevel imitation decision making(Bikhchandani et al., 1998). Thus, under informa-tion-based imitation, a prospective imitator willconsider characteristics of the competitor gener-ally, and of the new offering specifically, whencontemplating imitation.

Similarly, sociological approaches to informa-tion-based imitation have focused on “isomorphic

2010 1177Semadeni and Anderson

pressures,” based in uncertainty, to predict imita-tion. Under this perspective, imitation provides thebenefit not only of reducing uncertainty throughhomogeneity (DiMaggio & Powell, 1983; Haun-schild & Miner, 1997), but also of allowing a firm toeconomize on “search costs” by assuming compet-itors face similarly uncertain environmental condi-tions (Cyert & March, 1963). Commenting on this,Lieberman and Asaba noted: “In highly uncertainenvironments, where quick action is necessary, im-itating others becomes an attractive decision rule”(2006: 373). Contributing to the benefits of imita-tion under environmental uncertainty is the as-sumption that the legitimacy of imitation increasesas more and more firms imitate (Deephouse, 1996),despite increasing competitive pressures as thespecific market niche fills (Deephouse, 1999). Tosummarize the economic and sociological perspec-tives, information-based imitation can be thoughtto have two critical antecedents: perceived infor-mation asymmetry, manifested in the belief that acompetitor may possess superior market informa-tion, and environmental uncertainty. Furthermore,the decision to imitate a new offering is likely mul-tilevel in nature, dependent on characteristics ofboth the competitor and the offering.

Rivalry-based imitation. In contrast to the typeof imitation posited under economic and sociolog-ical theories, rivalry-based imitation centers on ef-fort to maintain competitive parity—to keep thecompetitive status quo. The foundation for rivalry-based imitation is the assumption of low informa-tion asymmetry; that is, the view that market actorsare likely to possess demonstrably similar knowl-edge about market conditions (Lieberman & Asaba,2006). Situations of low information asymmetry aretypically found in more certain environments,where the relationships between market actors arebetter established and the outcome of competitiveactions, including imitation, are well known (Baum& Korn, 1996; Gimeno & Chen, 1998; Gimeno &Woo, 1996). In this context, imitation is used moreas competitive tool driven by intraindustry compet-itive dynamics than as a mechanism to deal withperceived information asymmetry or environmen-tal uncertainty.

Lieberman and Asaba (2006) argued, however,that information- and rivalry-based theories of im-itation are not mutually exclusive and that firmsmay follow both theories in pursuing imitation.Furthermore, as Lieberman and Asaba noted, it istheoretically plausible, though less likely, to haverivalry-based imitation under conditions of envi-ronmental uncertainty. The critical differentiatingcharacteristic, however, in favoring one theory overthe other, is level of information asymmetry, with

information asymmetry high for information-basedand low for rivalry-based. Given that environmen-tal uncertainty and high information asymmetryfavor information-based imitation, the questionthen turns to identifying specific antecedents toimitation in this context.

Organization-Level Characteristics AffectingImitation of an Innovation

Competitor organizational innovativeness. Com-petitor organizational innovativeness is a firm-levelconstruct defined as a competitor’s history of intro-ducing innovative offerings over time. An organi-zation’s history is part of its reputation (Fombrun,1996; Fombrun & Shanley, 1990). A competitor’sreputation and pattern of historical behavior pro-vide information about the competitor to stake-holders and other market actors, particularly thosefirms contemplating imitation of the competitor’sinnovation (Bikhchandani et al., 1998). Thus, acompetitor’s history of consistently introducing in-novative new offerings to a market may signal aninnovation competency that overshadows the un-certainty and illegitimacy of an innovative newoffering (Ashforth & Gibbs, 1990; Suchman, 1995).To state this idea differently, though the viability ofan innovative new offering may be viewed as un-certain, its purveyor’s track record in innovationmay mitigate this uncertainty. In this way, compet-itor organizational innovativeness is a signal thatthe competitor, at the organizational level, pos-sesses superior information regarding the introduc-tion of innovative offerings. Under information-based imitation theory, this information superioritysignal by the competitor is likely to encourage theimitation of an innovation of that competitor, sothat:

Hypothesis 1. There is a positive relationshipbetween a competitor’s organizational innova-tiveness and the imitation of an innovation ofthat competitor.

Competitor offering relatedness. A competitor’soffering relatedness is a firm-level construct de-fined as the collective similarity of the competitor’sofferings. Provision of closely related products orservices may signal that the competitor has supe-rior information about a particular market (Bikh-chandani et al., 1998). Conversely, providing morediffuse or unrelated offerings may signal that thecompetitor lacks a substantive information advan-tage in any one market. The information superioritydemonstrated by the competitor’s offering related-ness might likely bolster its reputation (Haas &Hansen, 2007) and, like its track record for innova-


tiveness, decrease the uncertainty surrounding aninnovation (Abrahamson, 1996). It is important tonote here the distinction between the informationsuperiority signaled by competitor organizationalinnovativeness and the information superiority sig-naled by competitor offering relatedness. Competi-tor organizational innovativeness is marked by thecompetitor’s historical record of successfully intro-ducing innovative offerings over time. In this way,the information superiority signal is broadly based,encompassing all offerings of the competitor. Incontrast, competitor offering relatedness is definedas the collective similarity of its offerings. The in-formation superiority signaled in this case is inregard to knowledge of a particular market space.We would expect firms to be more likely to imitatea competitor’s innovation when that competitor isperceived to have specialized knowledge in a par-ticular market. Therefore:

Hypothesis 2. There is a positive relationshipbetween the relatedness of a competitor’s offer-ings and the imitation of an innovation of thatcompetitor.

Offering-Level Characteristics Affecting Imitationof an Innovation

Our key research question is how a firm parses,and what weight it assigns, the signals communi-cated through organization- and offering-level char-acteristics to determine whether or not imitation isan appropriate response to the introduction of aninnovation by a competitor. Thus, although organ-ization-level competitor characteristics are impor-tant factors in the imitation decision, so too areoffering characteristics (Bikhchandani et al., 1998).Indeed, firm decision makers contemplating imita-tion are not ultimately thinking about imitating thecompetitor per se; rather, they are contemplatingimitation of an innovation of the competitor. Im-portantly, the act of introducing an innovation tothe market represents a substantive commitment onbehalf of the innovator to pursue an opportunity ina new market space (Milgrom & Roberts, 1986;Miller, 1983). As such, the innovation represents acredible signal of the competitor’s belief about cur-rent market conditions and, more specifically, ofthe competitor’s perceived confidence in the mar-ket’s receptiveness to the innovation. Indeed,whether the innovation is radical or incremental,bringing forward a new product or service repre-sents a significant investment on the part of theinnovator (Ettlie, Bridges, & O’Keefe, 1984). Thus,prospective imitators are expected to parse the in-formation contained in the signal—the characteris-

tics of the innovation itself—carefully for insightinto the innovator’s perceived understanding of themarket environment (Milgrom & Roberts, 1986).

In their respective discussions of the norms ofrationality, Meyer and Rowan (1977) and Abraha-mson (1996) suggested that, as the innovativenessof an offering increases, the less reasonable thatparticular offering appears to the market. In otherwords, the more radical the innovation, the lessconfidence the market has in its perceived appro-priateness or, equivalently, the more confidencethat the innovation is poorly aligned with the cur-rent needs of the market. Conversely, an incremen-tal innovation tends to engender a higher degree ofmarket confidence in its congruence with marketexpectations. Thus, the innovativeness of a newoffering represents a signal, or proxy, of the degreeof “fit” between the offering and the market’s per-ception of its reasonableness: high levels of inno-vativeness indicate a poor perceived fit betweenmarket and innovation, and low levels of innova-tiveness represent a higher degree of fit (Abraham-son, 1996). Under rationality logic, imitating ahighly innovative offering represents a higher like-lihood of violating rationality norms; that is, byimitating the innovation, the imitator may appearout of sync with the market and less rational andmay even, in more extreme cases, actually imperilits organizational reputation (Abrahamson, 1996;Lieberman & Asaba, 2006).

Ultimately, this argument suggests a negative re-lationship between a competitor’s offering innova-tiveness—defined as the extent to which a newoffering introduces concepts and ideas previouslyunknown to the market—and imitation of that of-fering. At low levels of offering innovativeness,there is less risk of violating the norms of ration-ality, and imitation of the competitor’s less innova-tive offerings is less likely to jeopardize a firm’sreputation. As innovativeness increases, however,so does the uncertainty surrounding the rationalityof the innovation. Applying information-based im-itation theory, at higher levels of offering innova-tiveness, firms may doubt the efficacy of the offer-ing and thus may be more likely to believe theirinternal information as opposed to attributing in-formation superiority to the competitor (Lieberman& Asaba, 2006). In other words, the innovativenessof an offering decreases the information superioritysignal, causing firms to rely more on their internalinformation rather than attribute information supe-riority to the competitor (Bikhchandani et al.,1998). Conversely, at lower levels of offering inno-vativeness, the uncertainty surrounding the inno-vation is demonstrably lower, and those contem-plating imitation may be more likely to jettison


their internal information, having more confidencein the competitor’s interpretation of the market(Bikhchandani et al., 1992). To summarize, innova-tion is inherently risky and uncertain (Dougherty &Heller, 1994), and the greater the innovativeness ofa new service offering, the less likely firm decisionmakers contemplating imitation are to believe thatthe competitor possesses superior market knowl-edge; therefore:

Hypothesis 3. There is a negative relationshipbetween the innovativeness of a competitor’soffering and the imitation of that offering.

Interaction of Organization- andOffering-Level Characteristics

As suggested in the discussion of information-based imitation, the decision to imitate is likelymultilevel in nature, encompassing both organiza-tion- and offering-level characteristics. However,the interaction between organization- and offering-level characteristics may predict differing imitationoutcomes, particularly given our arguments thatorganization-level characteristics favor imitation,and an offering-level characteristic dampens it. Re-solving this apparent contradiction and arriving atan imitation decision, however, depends on theparticular configuration of characteristics, as differ-ent imitation outcomes emerge on the basis of therange of possible combinations of organization- andoffering-level factors. Stated differently, in makingthe imitation decision, a potential imitator mayface situations of either complementary or contra-dictory information superiority signals that jointlyaffect the imitation decision.

Competitor organizational innovativeness andoffering innovativeness. For this interaction, wepredict the highest level of imitation under condi-tions of high organizational innovativeness andlow offering innovativeness, as this context repre-sents the clearest signal that a competitor possessesinformation superiority. The competitor has a highdegree of demonstrated innovation competency,fostering a greater belief that it has superior infor-mation in regards to the introduction of innovativenew offerings. Similarly, the offering itself is onlyincremental in its innovativeness, which decreasesthe uncertainty of the innovation’s congruencewith current market conditions. It thereby also pro-motes a greater belief that the competitor possessessuperior market knowledge. Conversely, we expectthe lowest level of imitation under low competitororganizational innovativeness and high competitoroffering innovativeness. In this context, the com-petitor signals no information superiority of any

kind; it has no demonstrable innovation compe-tency, nor does the high level of offering innova-tiveness suggest that the competitor possesses afundamentally superior understanding of currentmarket conditions.

The remaining two configurations are lessstraightforward. Indeed, current theory provideslittle guidance about whether the information su-periority signal of the organizational characteristicsor that of the offering will prevail in determining ifimitation is an appropriate competitive response.In both of these contexts, the potential imitator isreceiving contradictory information superiority sig-nals—high competitor organizational innovative-ness suggesting information superiority, but alsohigh offering innovativeness, which calls into ques-tion congruence between the innovation and mar-ket expectations. Similarly, although low offeringinnovativeness may signal that the firm is more intune with the market, a correspondingly low levelof competitor organizational innovativeness signalsno demonstrable innovation competency. In theabsence of theory to direct prediction as to whichsignal will dominate, we can only assume that thetwo signals will offset each other; that is, a pureform of the interaction exists, in which case thelevel of imitation would not change. Thus, a com-plex interaction of competitor offering innovative-ness and competitor organizational innovativenessis predicted to result in differing levels of imitation,in line with the following hypothesis:

Hypothesis 4a. The innovativeness of a com-petitor’s offering and the competitor’s organi-zational innovativeness interact in such a waythat low offering innovativeness and high or-ganizational innovativeness result in thehighest level of imitation, and high offeringinnovativeness and low organizational inno-vativeness result in the lowest level ofimitation.

Competitor offering relatedness and offeringinnovativeness. Similar to the configurations of-fered in Hypothesis 4a, two configurations of com-plementary information superiority signals thatpredict the highest and lowest levels of imitation,respectively, are expected in the interaction ofcompetitor offering relatedness and offering inno-vativeness. The level of imitation should be at itsgreatest when competitor offering relatedness ishigh and offering innovativeness is low, as thisconfiguration represents the clearest overall signalof competitor information superiority. Conversely,under conditions of low competitor offering relat-edness, when the competitor signals no demonstra-ble competence in a particular market and signals


no demonstrably superior understanding of marketconditions by introducing a highly innovative newoffering, the level of imitation should be at its low-est. Again, as with Hypothesis 4a, the imitationoutcomes of those configurations that representconflicting information superiority signals are lessclear (high competitor offering relatedness/high of-fering innovativeness, and low competitor offeringrelatedness/low offering innovativeness), and inthe absence of theory to direct which signal willdominate, we again can only assume that the twosignals will offset one another. Thus, the precedingcomplex interaction is offered, in keeping with thefollowing hypothesis:

Hypothesis 4b. The innovativeness of a com-petitor’s offering and the competitor’s offeringrelatedness interact in such a way that lowoffering innovativeness and high offering relat-edness result in the highest level of imitation,and high offering innovativeness and low of-fering relatedness result in the lowest level ofimitation.

METHODOLOGY

Study Context

The majority of empirical imitation research hasfocused on product or commodity markets (e.g.,Asaba & Lieberman, 1999; Delios, Gaur, & Makino,2008; Gimeno, Hoskisson, Beal, & Wan, 2005; Rhee,Kim, & Han, 2006); instead, we focused on theservices sector of the economy for three key rea-sons. First, the patent protection available in prod-uct-markets imposes a barrier to direct imitation. Afirm may desire to imitate an innovation, but patentprotection makes such imitation legally difficultand often excessively expensive (Greve, 2003). Yetin the services sector, relatively little protectionexists for the intellectual property purveyed by ser-vice firms (Bowen & Ford, 2002), the result beingfew legal impediments to imitation. Second, it isextremely difficult to know a priori the value of aninnovation, particularly in the professional ser-vices sector. Specifically, customers in this sectormust ex ante select services and service providers,although the full value of the services will only beknown ex post (Maister, 1993). This unknownheightens uncertainty in the imitation decision. AsLieberman and Asaba noted, “Arguably, imitationprocesses are most interesting in environmentscharacterized by uncertainty or ambiguity [andwithout uncertainty imitation] is comparativelystraightforward and well understood” (2006: 366–367). Lastly, as Bowen and Ford (2002) noted,though service- and product-based firms share

some competitive similarities, there are significantdifferences in their strategic management, pertain-ing to delivery systems, customer relations, humancapital needs, and the production and consump-tion of offerings by customers (activities that aresimultaneous in the case of service-based firms).The prevalence of product-based research on imi-tation thus raises a generalizability concern whenapplying those results to service-based firms andwarrants studying different contexts to test imita-tion theories.

Professional Services Firms

For the purposes of this study, pending andregistered service marks for the 50 largest man-agement consulting firms (Kennedy Information,1999) from 1989 to 1999 were extracted from theU.S. Trademark database.1 As discussed previ-ously, information-based imitation is predicatedon both environmental uncertainty and high in-formation asymmetry, and for this reason, wechose the period from 1989 to 1999, because it wasmarked by explosive growth and change in theprofessional services sector. U.S. Census Bureaudata indicate that during 1987–2002, the 50 largestmanagement consulting firms experienced an 810percent growth in revenue, a 788 percent increasein payroll, and a 364 percent increase in the num-ber of employees. Comparatively, over the sameperiod and on average, firms in the United Statesexperienced a 121 percent increase in revenue, a 7percent increase in payroll, and a 19 percent de-crease in the average number of employees. Duringthe study period, the firms in the sample weredealing with extraordinary increases in profes-sional staff and office openings, as well as a dra-

1 Service marks are essentially trademarks attached toservices rather than tangible goods. For example, onlyProcter & Gamble can use the term “Crest” to sell tooth-paste, and only Lenovo can use the term “ThinkPad” tosell laptops, and these trademarks are assigned to tangi-ble products. Likewise, only Fidelity can use the term“Magellan” to sell mutual funds, and only General Mo-tors can use the term “Goodwrench” to sell automotiverepair services. According to the U.S. Patent and Trade-mark Office, the purpose of marks is to “identify anddistinguish” the products or services of a firm. Two prin-cipal benefits flow from identifying and distinguishingamong products and services. First, they encourage theproduction of quality goods that are known by theirtrademark. Second, consumer search costs are reduced.Unlike patents, trademarks and service marks may berenewed indefinitely as long as they have not becomegeneric.


matic increase in the number of client firms, allwhile increasing the number of services offered andthe number of markets served.

Additionally, beginning in 1989, all firms hadthe opportunity to file marks before they offeredthe goods or services in commerce. This was re-ferred to as an “intent-to-use filing” on thepremise that a firm intended to use the filed-formark but was not currently doing so. Thus, firmswith products or services under developmentcould file for mark protection in parallel withdevelopment, rather than wait until after theproduct or service was offered to the public. Fi-nally, prior to 1989, the 50 largest consultingfirms had filed only 21 service marks, but in thestudy period the number increased exponen-tially, underscoring the dramatic growth of ser-vices offered by the firms. All of these factorscombine to create a high degree of environmentaluncertainty and information asymmetry.

Data extracted from the U.S. Trademark databaseincluded the service mark itself, the filing date, theregistration date, and a description of the servicesto which the service mark pertained. This data setincludes all service marks filed by these firms re-gardless of whether the service mark registrationwas ultimately granted or not. To these data weadded location and descriptive data for each firmdrawn from the Directory of Management Consult-ants (Kennedy Information, 1999) and the Insite 2Intelligence Database. In all, 557 service marks arerecorded, with approximately 51 percent of themregistered.

Dependent Variable

The dependent variable in this analysis—imita-tion—describes the degree to which a firm introducesan offering based wholly or in part on an innovationby a competitor. Measuring imitation is particularlydifficult in the context of professional services firms,because services are intangible and, by their nature,difficult to describe. To file for a service mark, everyfirm must provide a concise yet parsimonious listingof the specific services that the mark covers. Thisdescription lies at the heart of the service mark pro-tection: the firm wishes to define its service mark asbroadly as possible, but the Trademark Office seeks toclearly define boundaries and limit the scope of themark. We analyzed the description of services usingkey consulting terms identified from Kennedy Infor-mation’s (1999) Directory of Management Consult-ants, combined with geographic terms (e.g., state, na-tional, international, multinational, global, etc.), aswell as the 1987 SIC divisions (e.g., agriculture, for-estry, fishing, mining, construction, manufacturing,

transportation, etc.).2 In total, 252 unique consulting,geographic, and industry terms were identified. Thekey terms used in the description of services for agiven mark were then compared to the key terms inevery other mark.3 We obtained the imitation variableusing network analysis (Borgatti, Everett, & Freeman,1999), by building an affiliation matrix with theBonacich normalization routine (Bonacich, 1972),which essentially evaluates the correlation betweenthe use of key terms in the description of services formark i with the key terms used in the description ofservices for mark j (Wasserman & Faust, 1994). Theformula for imitation was as follows:

Imitationij �(n11 � n22) � �n11 � n22 � n12 � n21

(n11 � n22) � (n12 � n21),

wheren11 � the number of key terms in common be-

tween mark i and mark j,n12 � the number key terms used by mark i

minus n11,n21 � the number key terms used by mark j

minus n11, andn22 � the total number of key terms used in all

marks minus n11 plus n12 plus n21.

This formula yields a value between 0 and 1,with 0 indicating no words in common and 1 indi-cating that all words included in the descriptionare common to both marks.

Independent Variables

Focal firm (firm j) offering innovativeness was anorder-of-entry variable that registered to what de-gree a mark filing used concepts and ideas intro-duced previously, versus new concepts and ideas.To obtain this measure, it was necessary to evaluatethe order of use of key management consultingterms: competitors that filed marks using terms ear-lier received higher scores than those that filedmarks using the terms later. The terms used hereare the same as those used to evaluate the imitationvariable. Additionally, since a given mark couldlist many management consulting terms, it was im-portant to score the order of use of each term. This

2 The key consulting terms have been expanded witheach edition of this directory, and to date, there havebeen no deletions from the list. The list is available fromthe authors upon request.

3 We expanded the key term list to capture variationsas well. For example, for the term “long-term planning,”we also added “long term planning,” so as not to miss thekey term because of the omission of a hyphen.


creates the relative innovativeness of the termsused. For example, if a competitor was the first touse term t, it received a 1. If it was the second touse the term, it received the value 1⁄2, and so forth.If the competitor did not use the given term, we setthe value to 0 to avoid a divide-by-zero error. Thescores for each of these terms were summed foreach of the marks. Summation yielded a raw scorethat we then divided by the number of words in themark’s description of services to normalize for thelength of the description of services. Normalizationprevented possible score inflation when marks hadlonger descriptions, and it yielded a weighted scorethat we then logged to normalize its distribution.

Next, competitor (firm i) organizational innova-tiveness measured the collective innovative historyof a firm’s offerings. We operationalized competitororganizational innovativeness as the average of allprior offering innovativeness scores for a givencompetitor, producing an approximation of thecompetitor’s past propensity to introduce innova-tive services over time, with higher values indicat-ing greater innovativeness.4

Finally, competitor (firm i) offering relatednessmeasured how closely related, or similar, the offer-ings of a competitor were to each other. To con-struct this variable, we took the average imitationamong the marks of the particular competitor. Inother words, this variable measured the degree towhich, on average, the services provided by a com-petitor were tightly clustered or diffuse. Essen-tially, calculation was similar to that used to createthe dependent variable but measured within-firm(rather than between-firm) imitation.

Control Variables

Several control variables were appropriate.First, given that many management consultingservices are subject to trends (Abrahamson, 1991,1996; Abrahamson & Rosenkopf, 1993), it wasimportant to control for the time between thefiling of a competitor’s mark and the filing of afocal firm’s mark. Because trends may peakquickly and then drop off or may grow exponen-tially over time, we controlled for nonlinear ef-fects as well, using elapsed time and elapsed timesquared.5 Second, we controlled for the priorimitation of the competitor by the focal firm(prior imitation ij), to mitigate the potential in-

fluence of rivalry-based imitation. This variablewas constructed by averaging the prior imitationof the competitor’s services by the focal firm.Third, we controlled for the quantity of priormark filings (total marks filed), as competitorsthat file a greater number of marks may have agreater likelihood of being imitated. Next, wecontrolled for the age and size of both competitorand firm. Prior research has shown that largerand more established competitors are more oftenimitated (Haunschild & Miner, 1997). Age wasthe number of years since a firm’s founding, andsize was the firm’s rank among the top 50 man-agement consulting firms in terms of revenue. Asthese two variables are collinear, both were cen-tered. Fifth, we controlled for the geographic lo-cation of a firm’s headquarters by adding a dummyvariable coded 1 if the firm was U.S.-based and 0 if itwas not. Approximately 20 percent of the firms in thesample were not headquartered in the United States.Sixth, we controlled for ownership structure, settingthe variable private to 1 if a firm was privately heldand to 0 if it was not. Seventh, for intent-to-use fil-ings, a variable was set to 1 if the mark was filed asintent-to-use and 0 if it was not. Finally, year dum-mies were added as controls for the effects of con-temporaneous correlation (Certo & Semadeni,2006). In addition to these control variables, a pro-spective imitator’s offering innovativeness and or-ganizational innovativeness were included.

Statistical Method

A random-effects time series cross-sectional regres-sion model was used to estimate the models (Greene,2003). A Hausman test indicated that the random-effects model was appropriate.6 The sample con-tained 557 unique observations over 11 years, yield-ing an unbalanced time series cross-sectional panel.Conversion of the data into dyadic pairs and temporalcorrection yielded 114,750 dyadic observations. As arobustness check, the models were also estimatedusing generalized least squares regression analysiscontrolling for heteroskedasticity, and the results

4 If the firm has only one service mark then, by defi-nition, organizational innovativeness was set to zero.

5 We transformed elapsed time to normalize itsdistribution.

6 The Hausman test evaluates the null hypothesis thatthe coefficients estimated by the efficient random-effectsestimator are the same as those estimated by the consis-tent fixed-effects estimator. If the test results are insignif-icant, then a random-effects model is appropriate andpreferred because it allows for both within and betweeninformation to calculate estimates; in contrast, the fixed-effects model only allows for within information to beused. Please see Certo and Semadeni (2006) for a moreextensive discussion of these two forms of modeling.


were essentially similar. All models were estimatedin Stata 11 (StataCorp, 2009).

RESULTS

Table 1 presents the summary statistics and cor-relations, and Table 2 presents the results of themodel. In the tables, firm i represents a competitorand firm j represents a focal firm. Examination ofthe correlation table indicated that some of thevariables have moderately high correlations, so toprotect against multicollinearity, we computedvariance inflation factors, and all were in the ac-ceptable range. To assess overall model fit, chi-square statistics and Akaike’s information criterion(AIC) were calculated and are reported.7

Hypothesis 1 predicts a positive relationship

between a competitor’s organizational innova-tiveness and imitation of its innovation. Hypoth-esis 2 predicts a positive relationship betweenthe competitor’s offering relatedness and imita-tion of the innovation. Model 2 in Table 2 teststhese hypotheses, and both are strongly sup-ported (� � 0.04, p � .001; � � 0.12, p � .001,respectively). Thus, the results suggest that firmswhose decision makers contemplate imitationlikely consider competitors with a history of in-novation and competitors with highly related of-ferings as having superior information about mar-ket conditions and that they are more likely toimitate an innovation of those competitors. Hy-pothesis 3 argues that offering innovativenesswill be negatively related to imitation of the in-novation. Again, model 2 in Table 2 tests thishypothesis, and it is strongly supported (� ��0.07, p � .001). Therefore, as hypothesized, theresults suggest that a prospective imitator likelydoubts the efficacy of highly innovative offerings

7 A decrease in the AIC indicates better model fit(Akaike, 1974).

TABLE 1Summary Statistics and Correlationsa

Variable Mean s.d. 1 2 3 4 5 6 7 8

1. Imitation 0.56 0.442. Elapsed time �0.01 7.35 �.01***3. Elapsed time

squared54.03 78.17 �.02*** .22***

4. Prior imitation ij 0.04 0.20 .05*** .01*** .02***5. Firm j total marks

filed28.16 38.47 .04*** �.03*** �.02*** �.14***

6. Firm j intent-to-use 0.55 0.50 .06*** .00 �.04*** �.05*** .09***7. Firm j age �0.96 37.17 .03*** .01* �.03*** �.14*** .37*** �.07***8. Firm j size �0.51 13.95 .01*** .01*** �.02*** �.17*** .46*** �.05*** .64***9. Firm j U.S.-based 0.94 0.23 �.01*** �.01*** �.02*** �.12*** .15*** .02*** .29*** .09***

10. Firm j private 0.72 0.45 .02*** .01** �.00 �.03*** .33*** �.10*** .36*** .37***11. Firm j offering

innovativeness�0.09 1.31 �.05*** .02*** .01* .07*** �.07*** �.01*** .08*** .01*

12. Firm j organizationalinnovativeness

�0.06 0.97 �.01** .05*** .01* �.00 .00 .08*** .18*** .00

13. Firm i total marksfiled

11.39 12.41 .01*** .03*** �.03*** �.12*** �.09*** �.01** �.05*** �.06***

14. Firm i intent-to-use 0.45 0.50 .03*** .02*** �.04*** �.03*** �.02*** �.01* �.01*** �.01***15. Firm i age �1.41 37.17 .06*** .04*** �.03*** �.12*** �.04*** �.00 �.05*** �.04***16. Firm i size �0.76 13.20 .05*** �.01* �.01* �.11*** �.07*** �.00 �.05*** �.06***17. Firm i U.S.-based 0.94 0.23 �.01*** .06*** .02*** �.17*** �.02*** �.00 �.01*** �.01***18. Firm i private 0.72 0.45 .03*** .01** .02*** �.04*** �.05*** .00 �.03*** �.04***19. Firm i offering

innovativeness0.01 1.00 �.05*** �.05*** �.02*** .03*** .01*** .00 .00 .01**

20. Firm i organizationalinnovativeness

0.03 1.03 .00 .01*** �.03*** .04*** .03*** �.00 .01*** .03***

21. Firm i offeringrelatedness

0.01 1.04 .09*** .04*** .04*** �.06*** .02*** .01 .01*** .01***

a Firm i: competitor; firm j: focal firm.* p � .05

** p � .01*** p � .001


and will tend to forgo imitation, trusting in theirown internal information superiority.8

Hypotheses 4a and 4b argue that competitor or-ganizational innovativeness and competitor offer-ing relatedness, respectively, will interact with of-fering innovativeness to predict differing imitationoutcomes depending on the configuration of theconstructs. This argument is based on the notionthat, in any given imitation decision, a prospectiveimitator may be faced with complementary or con-tradictory signals from organization- and offering-level characteristics. Testing these hypotheses re-quired two steps. In the first step, the interactionvariables were added to the research model andtested in models 3 and 4 for Hypotheses 4a and 4b,respectively, with both interactions highly signifi-cant (� � �0.04, p � .001; � � 0.04, p � .001). Inthe second step, point estimates of each interactionat one standard deviation from the mean were cal-

culated. A postestimation Wald test provided lin-ear comparisons of these point estimates and indi-cated that all comparisons were statisticallydifferent at the .001 level. These results are pre-sented and discussed in greater detail below.

Figure 1 shows a graph and point estimates forthe effects of the interaction of competitor organi-zational innovativeness and offering innovative-ness on imitation. As predicted, the highest andlowest levels of imitation occurred when the sig-nals communicated by the characteristics werecomplementary. The highest level of imitation oc-curred under conditions of high competitor organ-izational innovativeness and low offering innova-tiveness (the marginal effect being a 0.06 change inimitation), as this configuration represents theclearest signal that a competitor possesses superiorinformation about current market conditions. Incontrast, imitation was far lower when the compet-itor was signaling essentially no information supe-riority by introducing a highly innovative new of-fering and having no demonstrable organizationalinnovation competency (the marginal effect being a�0.04 change in imitation). The conditions under

8 Post hoc analysis of the coefficients revealed that thecompetitor organizational innovativeness, competitor of-fering relatedness, and offering innovativeness resultsare statistically different from one another.

TABLE 1Continued

9 10 11 12 13 14 15 16 17 18 19 20

.09***�.02*** .04***

�.04*** .05*** .43***

�.02*** �.05*** �.01* �.01**

�.00 �.01** �.00 �.01* .18***�.01*** �.03*** �.01* �.01** .35*** .06***�.01*** �.04*** .00 .01* .37*** �.02*** .59***�.01*** �.01*** �.00 �.00 .20*** .65*** .25*** .16***�.01*** �.05*** �.00 �.00 .32*** �.11*** .38*** .43*** �.04***

.00 .01* �.00 �.01 �.20*** �.04*** .02*** �.09*** �.10*** .05***

.01* .02*** �.01** �.02*** �.07*** .07*** .06*** �.33*** �.27*** .04*** .39***

.00 .01** .01* .01* �.06*** �.05*** �.02*** .07*** .35*** �.12*** �.09*** �.27***


which the signals were contradictory produced no-ticeably intriguing results. Under conditions of lowoffering innovativeness and low competitor organ-izational innovativeness, the level of imitation wasessentially neutral (no appreciable marginal changein imitation). However, under conditions of highcompetitor organizational innovativeness (a posi-tive information superiority signal) and high offer-ing innovativeness (a negative information superi-ority signal), the level of imitation fell (the marginaleffect being a �0.02 change in imitation). This pat-tern suggests that although a competitor may have ademonstrable organizational competency for inno-vation, such a competency does not overshadowthe inherent uncertainty, and thereby the perceivedirrationality, of a highly innovative offering.

Figure 2 shows a graph and point estimates forthe interaction of competitor offering relatedness

and offering innovativeness on imitation. As withFigure 1, the highest and lowest levels of imitationoccurred under conditions of complementary infor-mation superiority signals (marginal effects yield-ing 0.06 and �0.09 changes, respectively). Also inkeeping with Hypothesis 4b, and similarly to thefindings reported in Figure 1, there was essentiallyno change in imitation under conditions of lowcompetitor offering relatedness and low offeringinnovativeness. However, imitation increased un-der conditions of high competitor offering related-ness (a positive signal) and high offering innova-tiveness (a negative signal), with a marginal effectbeing a 0.03 change in imitation. In this context, thestrength of the positive organization-level character-istic outweighs the influence of the negative offering-level characteristic, so that competitors with a de-monstrable competence in a particular market signal

TABLE 2Results of Random-Effects Regression Analysis for Imitationa

Variableb Model 1 Model 2 Model 3 Model 4 Model 5

Intercept 0.76*** (0.09) 0.84*** (0.09) 0.86*** (0.09) 0.85*** (0.09) 0.86*** (0.09)Elapsed time �0.01*** (0.00) �0.02*** (0.00) �0.02*** (0.00) �0.02*** (0.00) �0.02*** (0.00)Elapsed time squared �0.01*** (0.00) �0.02*** (0.00) �0.02*** (0.00) �0.01*** (0.00) �0.01*** (0.00)Prior imitation ij 0.06*** (0.01) 0.06*** (0.01) 0.06*** (0.01) 0.06*** (0.01) 0.06*** (0.01)Firm j total marks filed 0.07*** (0.00) 0.07*** (0.00) 0.07*** (0.00) 0.07*** (0.00) 0.07*** (0.00)Firm j intent-to-use 0.04*** (0.00) 0.04*** (0.00) 0.04*** (0.00) 0.04*** (0.00) 0.04*** (0.00)Firm j age 0.13*** (0.00) 0.13*** (0.00) 0.13*** (0.00) 0.13*** (0.00) 0.13*** (0.00)Firm j size �0.01 (0.00) �0.01 (0.00) �0.01 (0.00) �0.01 (0.00) �0.01 (0.00)Firm j private 0.01 (0.06) 0.01 (0.06) 0.01 (0.06) 0.01 (0.06) 0.01 (0.06)Firm j U.S.-based �0.02 (0.07) �0.02 (0.07) �0.02 (0.07) �0.02 (0.07) �0.02 (0.07)Firm j offering innovativeness �0.06*** (0.00) �0.06*** (0.00) �0.06*** (0.00) �0.06*** (0.00) �0.06*** (0.00)Firm j organizational

innovativeness�0.05 (0.02) �0.06 (0.02) �0.05 (0.02) �0.05 (0.02) �0.05 (0.02)

Firm j total marks filed �0.00 (0.00) �0.01** (0.00) �0.01*** (0.00) �0.01*** (0.00) �0.01*** (0.00)Firm i intent-to-use 0.04*** (0.00) 0.04*** (0.00) 0.04*** (0.00) 0.04*** (0.00) 0.04*** (0.00)Firm i age 0.07*** (0.00) 0.07*** (0.00) 0.07*** (0.00) 0.07*** (0.00) 0.07*** (0.00)Firm i size 0.05*** (0.00) 0.04*** (0.00) 0.04*** (0.00) 0.04*** (0.00) 0.04*** (0.00)Firm i private �0.01*** (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00) 0.00 (0.00)Firm i U.S.-based �0.03*** (0.01) �0.07*** (0.01) �0.07*** (0.01) �0.07*** (0.01) �0.07*** (0.01)Firm i offering innovativeness �0.07*** (0.00) �0.07*** (0.00) �0.07*** (0.00) �0.07*** (0.00)Firm i organizational

innovativeness0.04*** (0.00) 0.05*** (0.00) 0.05*** (0.00) 0.05*** (0.00)

Firm i offering relatedness 0.12*** (0.00) 0.12*** (0.00) 0.11*** (0.00) 0.11*** (0.00)Firm i organizational

innovativeness � firm ioffering innovativeness

�0.04*** (0.00) �0.02*** (0.00)

Firm i offering relatedness �firm i offeringinnovativeness

0.04*** (0.00) 0.03*** (0.00)

Log-likelihood �63,433.99 �62,462.17 �62,403.81 �62,384.27 �62,362.8�2 2,543.26 4,486.9 4,603.61 4,642.71 4,685.64AIC 126,927.97 124,990.33 124,875.63 124,836.53 124,795.6AIC improvement 1,937.64 114.70 39.10 40.93

a n � 114,750. Standardized coefficients are reported, with standard errors in parentheses. Year dummies were included in analyses butare not reported here.

b Firm i: competitor; firm j: focal firm.** p � .01

*** p � .001


sufficient information superiority to eclipse the un-certainty of even the most innovative offerings.

The analysis yields several other noteworthy re-sults. First, elapsed time and elapsed time squaredare both negative. This indicates that, as timeelapsed, the level of imitation decreased at an in-creasing rate. We take this to suggest that from theirintroduction, the services purveyed by these firmsrapidly became obsolete, decreasing their imita-tion. An alternative explanation for this finding isthe possibility that an innovation was very rapidlyimitated by a significant number of firms in thesample. In this scenario, such rapid imitation effec-tively results in market saturation, whereby anyremaining firms that have not imitated the innova-tion are far less likely to introduce a substantiallysimilar service in such a crowded market space(Makadok, 1998). Next, average prior imitation ispositive and significant, suggesting that past imita-tion is a good predictor of future imitation. As weelaborate in the next section, this variable is impor-tant to controlling for rivalry-based theories of im-itation. Finally, U.S.-based competitors were imi-tated less than non-U.S.-based competitors. We areuncertain how to interpret this finding other thanto say it warrants further examination.

We conducted several tests to evaluate the robust-

ness of our results. First, to ensure that the initialmark filing of either a competitor or a focal firm didnot bias our results, particularly if the firm had one oronly a few filings, we dropped the first mark filing foreach firm, and the results held. Next, we wished toknow if the ultimate registration of a mark influencedour results. To test for this, we included a dummyvariable that indicated if the mark was abandonedprior to registration, and once again, the results held.Third, 21 cases were lost in the sample by our limit-ing the study period to 1989–99. To evaluate if thiscensoring influenced our results, we included those21 marks in our analysis, and the results did not differsignificantly. Finally, we provide in model 5 of Table2 the joint estimation of the two interactions, and theresults are essentially the same as the models inwhich the two interactions are estimated separately.In sum, these robustness checks enhance our confi-dence in the reported results.

DISCUSSION

Organization- and Offering-Level Predictorsof Imitation

Although the ubiquitous investment disclosurewarns that “past performance is no guarantee of

FIGURE 1Moderation of Competitor Organizational Innovativeness and Imitation by Offering Innovativeness


future results,” there is evidence that a competitor’sorganizational innovativeness, or its history of in-troducing innovative offerings over time, increasesimitation of that competitor’s innovations. In keep-ing with Ashforth and Gibbs (1990) and Suchman(1995), the competitor’s organizational innovative-ness signals an innovation competency to thosecontemplating imitation, in such a way that thiscompetency conveys legitimacy to new offerings.Similarly, competitor offering relatedness is posi-tively related to imitation. Highly related offeringssignal greater demonstrable competency in a par-ticular market area than more disparate offerings(Calantone & Cooper, 1981). Like competitor organ-izational innovativeness, competitor offering relat-edness likely signals perceived information superi-ority, which increases imitation.

Organization-level factors resulted in increasedimitation, yet the innovativeness of an offering it-self dampened imitation. As a robustness check forthis result, we conducted post hoc analysis to testfor the possibility of a curvilinear relationship be-tween an offering’s innovativeness and its imita-tion—examining the possibility that extreme in-novation deterred imitation but that moderateinnovation encouraged it—and the results were notsignificant. Thus, at its most basic level, the moreinnovative a competitor’s offering, the less likely afirm is to imitate that innovation. This finding sup-

ports the argument based on norms of rationality atthe offering level; specifically, radical innovationsmay violate market expectations of reasonableness(Abrahamson, 1996). Applying information-basedimitation theory, a firm may be more likely to be-lieve that the competitor has “guessed wrong” byintroducing such a radical innovation, and it istherefore unlikely to imitate the innovation (Lieber-man & Asaba, 2006).

It is important to discuss here the role of re-sources and capabilities in imitation decisions.Whenever imitation occurs among competitorfirms, there is the possibility of faddish behaviorand “bandwagons” (Abrahamson, 1991, 1996;Abrahamson & Rosenkopf, 1993; Palley, 1995).Such behavior, however, can be detrimental tothose market actors that are quick to imitate butlack the concomitant understanding needed to ac-tually make the product or service work (Lieber-man & Asaba, 2006). This statement raises the ques-tion of the role of capabilities in the making ofimitation decisions. Arguably, an imitation deci-sion has two stages. First, those contemplating im-itation must determine if their firm should imitate;that is, they must determine the appropriateness ofimitation as a competitive response. Second, theprospective imitators must determine whether theycould imitate; that is, evaluate whether the firmpossesses, or could acquire, the resources and ca-

FIGURE 2Moderation of Competitor Offering Relatedness and Imitation by Offering Innovativeness


pabilities necessary to implement the imitation de-cision. Notably, information-based theories of imi-tation are fundamentally concerned with the firststage of the imitation decision, or whether a firmshould or should not imitate an innovation be-cause, as Lieberman and Asaba (2006) noted,firms may actually pursue imitation regardless ofwhether they actually possess the necessary ca-pabilities to implement the imitation (implemen-tation is here defined as the profitable introduc-tion of an offering that imitates an earlier offeringby a competitor).

Both economics- and sociology-based informa-tion theories of imitation suggest cases in which animitation is pursued regardless of an inherent ca-pability to actually implement the imitation. In theeconomics literature, Palley (1995) and Scharfsteinand Stein (1990) wrote that inferior managers mayknowingly imitate managers perceived to be supe-rior solely to bolster their own reputations, regard-less of the potential consequences of their actions.Sociological research has identified similar behav-ior among firms that frequently engage in late-stageimitation—such firms are often not concerned withimplementation per se; rather, they engage in imi-tation, particularly of more prestigious firms, in anattempt to increase their market legitimacy (Flig-stein, 1985; Lieberman & Asaba, 2006). Further-more, a lack of consideration of inherent capabilitymay be especially salient in contexts in which re-sources and capabilities are fundamentally similaramong market actors (Lieberman & Asaba, 2006),such as the management consulting industry. Insuch a context, firms tend to possess demonstrablysimilar resource stocks (i.e., consultants with grad-uate business degrees), and their capabilities areinherently fungible. Collectively, these notions sug-gest that the findings in the current study regardingimitation decision making based on perceived infor-mation superiority signaled by organization- and of-fering-level characteristics are particularly salient, asthese firms’ decision makers are likely to assume thatthey already possess the necessary capabilities to im-plement an imitation.

Nonetheless, the possibility exists that a prospec-tive imitator’s internal capabilities could confoundthe imitation decision as investigated in the currentstudy. Although we did not capture a variable di-rectly measuring the internal capabilities of a pro-spective imitator, the intent-to-use variable, used asa control in the analysis, may represent a rough, butappropriate, proxy for firm capability. As discussedpreviously, a firm makes an intent-to-use filing toexpress a desire to provide a particular service tothe market that it is not currently offering. Such afiling thus suggests lack of the capabilities neces-

sary to offer the service to the market at the time thefiling is made. As Table 2 shows, the firm j (pro-spective imitator) intent-to-use dummy variable(with 1 indicating an intent-to-use filing and 0 in-dicating no filing) is positive and significant acrossthe models. This finding indicates that a firm notcurrently offering a service, but with a desire to doso, is more likely to pursue imitation. A possibleexplanation for this result is that firms lacking thecapability to engage in imitation still pursue it in anattempt to balance legitimacy against competitiveforces (see Semadeni, 2006). Yet even in the pres-ence of such a capability consideration, the modelresults maintain our finding that perceived infor-mation superiority—or lack thereof—is likely apredominate antecedent to imitation. As a furtherrobustness test, we investigated the possibility thatfirm capabilities have an interactive effect on re-sults (i.e., having the capability, as evidenced bythe intent-to-use variable being set to zero, differ-entially affects the imitation decision). We testedthis idea by splitting the sample into two groupsbased on whether an intent-to-use filing was or wasnot made, and then running our hypothesized mod-els again. Neither the sign nor the significance ofthe coefficients of our variables of interest changedacross the two groups, providing further evidenceof the robustness of our findings.9

Interaction of Organization- andOffering-Level Characteristics

Hypotheses 4a and 4b state that a complex inter-action of the two organization-level characteristicswith offering innovativeness will result in signifi-cantly different imitation outcomes, and both ofthese hypotheses were supported. Notably, currenttheory was not informative in predicting imitationwhen information superiority signals were contra-dictory. In such cases, a pure form of the interac-tion would suggest that the signals carry equalweight, and therefore, should cancel each otherout, resulting in no change in imitation. In twoconfigurations, this was indeed the outcome. Whencompetitor organizational innovativeness was low(a negative signal), when competitor offering relat-edness was low (a negative signal), and when offer-ing innovativeness was also low for both of theseconditions (a positive signal), predicted imitationessentially dropped to zero, meaning that the signalweights were essentially equivalent. Such equiva-lency may be tied to the notion that incrementalinnovations tend to result in only marginal eco-

9 Results are available from the authors upon request.


nomic returns (i.e., the concept of low-risk, low-reward [see Milgrom & Roberts, 1986]). In this con-text, there may be little to be gained from imitatingthe innovation, and when that condition is com-bined with the lack of a clear information superi-ority signal at the organizational level, firms’ deci-sion makers seem to be somewhat ambivalent aboutimitating an only marginally innovative offering.

In the remaining two configurations of contradic-tory information superiority signals, we found evi-dence of dramatically different imitation behavior.Under conditions of high competitor organizationalinnovativeness (a positive signal) and high offeringinnovativeness (a negative signal), the overall levelof imitation decreased. Thus, the weight of theoffering-level characteristic was greater than that ofthe organization-level characteristic in this context.By contrast, under conditions of high competitoroffering relatedness (a positive signal) and highoffering innovativeness (a negative signal), imita-tion increased. In this context, the weight of theorganization-level characteristic eclipsed that ofthe offering-level characteristic. Given that the con-stant in both of these interactions is offering inno-vativeness, prospective imitators are therefore plac-ing differing weights on the information superioritysignals sent by the two organization-level charac-teristics. Indeed, as the results in Table 2 indicate,the coefficient size of competitor offering related-ness was consistently stronger in the models thanthat of competitor organizational innovativenessand offering innovativeness.

These findings offer three key insights. The firstis that, though competitors with a broad innovationcompetency may be more likely to be imitated,when such competitors introduce highly innova-tive offerings, the information superiority con-veyed by organizational innovativeness is not suf-ficient to overcome the uncertainty of a radicalinnovation. In short, firms will follow a competitorwith a track record for innovation, but not to thepoint of jeopardizing their own reputations andmarket legitimacy by imitating a dramatically inno-vative offering. The second is that firms seem muchmore comfortable imitating the innovations ofthose competitors with a demonstrable competencyin a particular market area. In these cases, firmsseem quick to jettison their own internal informa-tion in favor of following the actions of those com-petitors who are perceived to have superior knowl-edge about a specific market, regardless of whetherthe offering introduced by such competitors is rad-ical or incremental. Lastly, these results also pro-vide strong evidence that firms separate offeringfrom firm and firm from offering when making im-itation decisions or, in keeping with the observa-

tion of Chamberlain (1933), that what distinguishesa firm is different from what distinguishes theproducts of the firm.

Implications

In addition to making a theoretical contributionby identifying the relative weightings of organiza-tion- and offering-level information superiority sig-nals depending on context, our study’s findingsoffer two additional implications. First, there isevidence that under conditions of environmentaluncertainty, firms actively seek to balance the com-peting norms of rationality and progressiveness.Although this may be taken as empirical evidencesupporting previous theoretical work (e.g., Abra-hamson, 1996; Lieberman & Asaba, 2006), two newinsights emerge here. The first is that concern overviolating rationality appears to sufficiently dampenan appetite for imitating radical innovations. Inkeeping with Abrahamson (1996), firms’ decisionmakers are likely to view a competitor’s radicalinnovations with skepticism and to be unwilling torisk their reputation by imitating such radical newofferings. However, imitation increases at lowerlevels of offering innovativeness, underscoring thepower of norms of progressiveness. Thus, firmsmay not be willing to imperil their reputation byimitating radical innovations, yet they do seek tokeep pace with their competitive peers by imitatingrelatively more incremental innovations. Stateddifferently, we find empirical evidence that imita-tion is likely a mechanism through which firmsseek to appear forward-looking (under norms ofprogressiveness) by imitating less risky innova-tions, but they will avoid imitating innovations thatmake them appear extreme (under normsof rationality).

Second, inasmuch as higher offering innovative-ness resulted in a lower imitation, radical innova-tion appears to be a significant driver of differenti-ation in the services sector. From a competitivepositioning perspective, significant imitation oc-curs at lower levels of offering innovativeness, sug-gesting that both information- and rivalry-basedtheories of imitation may be driving imitation de-cisions. Rivalry theories suggest that imitation isbased on maintaining competitive parity, in such away that firms’ decision makers seek to diminish apotential competitive advantage through imitation(Lieberman & Asaba, 2006). Importantly, rivalry-based imitation is likely to occur in more certainenvironments and when information disparity be-tween market actors is low. At low levels of offeringinnovativeness, information asymmetry is dimin-ished, and firms may use imitation as a mechanism


to blur differentiation between themselves andtheir competitors—to maintain the competitivestatus quo. However, at high levels of offeringinnovativeness, perceived information asymme-try increases, thereby confounding imitation de-cisions. Thus, differentiation through radical in-novation is more achievable in the professionalservices sector, as rivals are far less likely toimitate such innovations.

Research Limitations and Directions forFuture Research

Like all social science research, this study doeshave certain limitations. First, as Lieberman andAsaba (2006) noted, there is the possibility of noimitation when two firms have similar offerings;rather, the two firms may have responded similarlyto the same environmental stimuli and introduceddemonstrably similar innovations concurrently.Although this occurrence is theoretically possible,Lieberman and Asaba (2006) also pointed out that itis highly unlikely to be prevalent. Given the exten-sively longitudinal nature of the data in the currentstudy, we feel confident that, if present, the effectof such occurrences is negligible.

Second, empirically distinguishing between ri-valry-based and information-based theories of imi-tation is difficult. Empirically, the inclusion ofprior imitation does control, to some degree, forrivalry between competitor and focal firm. Thus,we feel confident that the conclusions drawn fromthis research regarding the dominance of informa-tion-based theories of imitation in such a contextare well founded. Future research, perhaps quali-tative in nature, might explore individual imitationdecisions to unravel the information- and rivalry-based antecedents. The ability to separate andmodel rivalry and information factors in an imita-tion decision will likely provide helpful new in-sights into the imitation phenomenon.

Third, as with all empirical research, differentstudy contexts and different operationalizations ofconstructs have the potential to produce differentresults. The management consulting industry dur-ing 1989–99 was marked by dramatic change andgrowth and so was an appropriate context in whichto explore information-based imitation theory.Nonetheless, future research into different indus-tries and time periods is necessary to better estab-lish the generalizability of this study’s findings.

Lastly, because the management consulting in-dustry is secretive about its operations, we couldnot examine the performance of the offerings thesample firms introduced. Thus, the limitation isthat data were not available to control for a given

mark’s success or failure in the market. As a result,somewhat of a survivor bias is inherent to thestudy; that is, there is an assumption, particularlyin the construct of competitor organizational inno-vativeness, that at least a plurality of the competi-tor’s innovations were successful over time, or thecompetitor itself would become imperiled and fal-ter. If the data could be obtained, future researchinto the performance consequences of imitation de-cisions would begin to fill a critical gap in theimitation literature.

Conclusion

The purpose of this study was finer-grained ex-planation of the factors that predict imitation in animportant, yet underexplored, environmental con-text. This research identified both organization-and offering-level characteristics that influenceimitation, and more significantly, identified an in-teraction between these characteristics whereinlevel of imitation differs substantially dependingon the specific configuration of organization- andoffering-level factors. These findings have enrichedknowledge of the contributors to imitation decisionmaking and facilitate better understanding of howfirms resolve the follower’s dilemma—to imitateor not.

REFERENCES

Abrahamson, E. 1991. Managerial fads and fashions: Thediffusion and rejection of innovations. Academy ofManagement Review, 16: 586–612.

Abrahamson, E. 1996. Management fashion. Academy ofManagement Review, 21: 254–285.

Abrahamson, E., & Rosenkopf, L. 1993. Institutional andcompetitive bandwagons. Academy of Manage-ment Review, 18: 487–517.

Akaike, H. 1974. A new look at the statistical modelidentification. IEEE Transactions on AutomaticControl, 19: 716–723.

Asaba, S., & Lieberman, M. B. 1999. Why do firms be-have similarly? A study on new product introduc-tions in the Japanese soft-drink industry. Academyof Management Best Paper Proceedings, BPS: 01-08.

Ashforth, B. E., & Gibbs, B. W. 1990. The double-edge oforganizational legitimation. Organization Science,1: 177–194.

Baum, J. A. C., & Korn, H. J., 1996. Competitive dynamicsof interfirm rivalry. Academy of Management Jour-nal, 39: 255–291.

Bikhchandani, S., Hirshleifer, D., & Welch, I. 1992. Atheory of fads, fashion, custom, and cultural change


as information cascades. Journal of Political Econ-omy, 100: 992–1026.

Bikhchandani, S., Hirshleifer, D., & Welch, I. 1998.Learning from the behavior of others: Conformity,fads, and informational cascades. Journal of Eco-nomic Perspectives, 12(3): 151–170.

Bonacich, P. 1972. Technique for analyzing overlappingmemberships. In H. Costner (Ed.), Sociologicalmethodology: 176–185. San Francisco: Jossey-Bass.

Borgatti, S. P., Everett, M. G., & Freeman, L. C. 1999.Ucinet 5 for Windows (v. 5.4). Natick, MA: AnalyticTechnologies.

Boulding, W., & Christen, M. 2008. Disentangling pio-neering cost advantages and disadvantages. Market-ing Science 27: 699–716.

Bowen, J., & Ford, R. C. 2002. Managing service organi-zations: Does having a “thing” make a difference?Journal of Management, 28: 447–469.

Calantone, R., & Cooper, R. G. 1981. New product sce-narios: Recipes for success. Journal of Marketing,45(2): 48–60.

Certo, S. T., & Semadeni, M. 2006. Strategy research andpanel data: Evidence and implications. Journal ofManagement, 32: 449–471.

Chamberlin, E. 1933. The theory of monopolistic com-petition. Cambridge, MA: Harvard University Press.

Chen, M.-J. 1996. Technological expansion: The interac-tion between diversification strategy and organiza-tional capability. Journal of Management Studies,33: 649–666.

Chen, M.-J., & MacMillan, I. C. 1992. Nonresponse anddelayed response to competitive moves: The roles ofcompetitor dependence and action irreversibility.Academy of Management Journal, 35: 539–570.

Chen, M.-J., Su, K.-H., & Tsai, W. 2007. Competitivetension: The awareness-motivation-capability per-spective. Academy of Management Journal, 50:101–118.

Christensen, C. M. 1997. The innovator’s dilemma. Bos-ton: Harvard Business School Press.

Covin, J. G., & Slevin, D. P. 1991. A conceptual model ofentrepreneurship as firm behavior. Entrepreneur-ship: Theory and Practice, 16(1): 7–24.

Cyert, R. M., & March, J. G. 1963. A behavioral theory ofthe firm. Englewood Cliffs, NJ: Prentice-Hall.

Deephouse, D. L. 1996. Does isomorphism legitimate?Academy of Management Journal, 39: 1024–1039.

Deephouse, D. L. 1999. To be different, or to be the same?It’s a question (and theory) of strategic balance. Stra-tegic Management Journal, 20: 147–166.

Delios, A., Gaur, A. S., & Makino, S. 2008. The timing ofinternational expansion: Information, rivalry andimitation among Japanese firms, 1980–2002. Journalof Management Studies, 45: 169–195.

DiMaggio, P. J., & Powell, W. W. 1983. The iron cagerevisited: Institutional isomorphism and collectiverationality in organizational fields. American Socio-logical Review, 48: 147–160.

Dougherty, D., & Heller, T. 1994. The illegitimacy ofsuccessful product innovation in established firms.Organization Science, 5: 200–218.

Ettlie, J. E., Bridges, W. P., & O’Keefe, R. D. 1984. Orga-nization strategy and structural differences for radi-cal versus incremental innovation. ManagementScience, 30: 682–695.

Ferrier, W. J. 2001. Navigating the competitive land-scape: The drivers and consequences of competitiveaggressiveness. Academy of Management Journal,44: 858–877.

Fligstein, N. 1985. The spread of the multidivisionalform among large firms, 1919–1979. American So-ciological Review, 50: 377–391.

Fombrun, C. 1996. Reputation: Realizing value from thecorporate image. Boston: Harvard Business SchoolPress.

Fombrun, C., & Shanley, M. 1990. What’s in a name?Reputation building and corporate strategy. Acad-emy of Management Journal, 33: 233–258.

Gimeno, J. 1999. Reciprocal threats in multimarket ri-valry: Staking out “spheres of influence” in the U.S.airlines industry. Strategic Management Journal,20: 101–128.

Gimeno, J., & Chen, M.-J. 1998. The dynamics of compet-itive positioning: A pair-wise perspective. Academyof Management Best Paper Proceedings, BPS:01–08.

Gimeno J., Hoskisson R. E., Beal, B. D., & Wan W. P. 2005.Explaining the clustering of international expansionmoves: A critical test in the U.S. telecommunica-tions industry. Academy of Management Journal,48: 297–319.

Gimeno, J., & Woo, C. Y. 1996. Hypercompetition in amultimarket environment: The role of strategic sim-ilarity and multimarket contact in competitive dee-scalation. Organization Science, 7: 322–341.

Greene, W. H. 2003. Econometric analysis. Upper Sad-dle River, NJ: Prentice Hall.

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

Greve, H. R. 2003. A behavioral theory of R&D expendi-tures and innovations: Evidence from shipbuilding.Academy of Management Journal, 46: 685–702.

Guler, I., Guillen, M. F., & MacPherson, J. M. 2002. Globalcompetition, institutions, and the diffusion of organ-izational practices: The international spread of theISO 9000 quality certificates. Administrative Sci-ence Quarterly, 47: 207–232.


Haas, M. R., & Hansen, M. T. 2007. Different knowledge,different benefits: Toward a productivity perspectiveon knowledge sharing in organizations. StrategicManagement Journal, 28: 1133–1153.

Haunschild, P. R. 1993. Interorganizational imitation:The impact of interlocks on corporate acquisitionactivity. Administrative Science Quarterly, 38:564–592.

Haunschild, P. R., & Miner, A. S. 1997. Modes of interor-ganizational imitation: The effects of outcome sa-lience and uncertainty. Administrative ScienceQuarterly, 41: 472–500.

Kennedy Information. 1999. Directory of ManagementConsultants (9th ed.). Fitzwilliam, NH: KennedyInformation.

Lambkin, M. 1988. Order of entry and performance innew markets. Strategic Management Journal,9(S1): 127–140.

Lee, P. 2001. What’s in a name.com? The effects of“.com” name changes on stock prices and tradingactivity. Strategic Management Journal, 22: 793–804.

Lee, J., Lee, K., & Rho, S. 2002. An evolutionary perspec-tive on strategic group emergence: A genetic algo-rithm-based model. Strategic Management Journal,23: 727–746.

Levinthal, D. A., & March, J. G. 1993. The myopia oflearning. Strategic Management Journal, 14(spe-cial issue): 95–112.

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

Lieberman, M. B., & Montgomery, D. B. 1988. First-moveradvantages. Strategic Management Journal, 9(spe-cial issue): 41–58.

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

Maister, D. H. 1993. Managing the professional servicefirm. New York: Free Press.

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

March, J. G. 1991. Exploration and exploitation in organ-izational learning. Organization Science, 2: 71–87.

Meyer, J., & Rowan, B. 1977. Institutionalized organiza-tions: Formal structure as myth and ceremony.American Journal of Sociology, 83: 340–363.

Milgrom, P., & Roberts, J. 1986. Price and advertisingsignals of product quality. Journal of Political Econ-omy, 78: 311–329.

Miller, D. 1983. The correlates of entrepreneurship in

three types of firms. Management Science, 29: 770–791.

Palley, T. I. 1995. Safety in numbers: A model of mana-gerial herd behavior. Journal of Economic Behaviorand Organization, 28: 443–450.

Porter, M. E. 1980. Competitive strategy: Techniquesfor analyzing industries and competitors. NewYork: Free Press.

Rhee, M., Kim, Y.-C., & Han, J. 2006. Confidence inimitation: Niche-width strategy in the U.K. automo-bile industry. Management Science, 52: 501–513.

Scharfstein, D. S., & Stein, J. C. 1990. Herd behavior andinvestment. American Economic Review, 80: 465–479.

Semadeni, M. 2006. Minding your distance: How man-agement consulting firms use service marks to posi-tion competitively. Strategic Management Journal,27: 169–187.

StataCorp. 2009. Stata statistical software: Release 11.College Station, TX: StataCorp.

Suarez, F. F., & Lanzolla, G. 2007. The role of environ-mental dynamics in building a first mover advantagetheory. Academy of Management Review, 32: 377–392.

Suchman, M. C. 1995. Managing legitimacy: Strategicand institutional approaches. Academy of Manage-ment Review, 20: 571–610.

Terlaak, A., & King, A. A. 2007. Follow the small? Infor-mation revealing adoption bandwagons when ob-servers expect larger firms to benefit more fromadoption. Strategic Management Journal, 28:1167–1185.

Wasserman, S., & Faust, K. 1994. Social network anal-ysis: Methods and applications. New York: Cam-bridge University Press.

Williamson, O. E. 1983. Credible commitments: Usinghostages to support exchange. American EconomicReview, 73: 519–540.

Matthew Semadeni ([email protected]) is an assis-tant professor of strategic management in the KelleySchool of Business at Indiana University. He received hisPh.D. from Texas A&M University. His research interestsinclude competitive strategy, top management teams,and innovation/knowledge creation.

Brian S. Anderson ([email protected]) is a visiting assis-tant professor of strategic management in the KelleySchool of Business at Indiana University. He received hisPh.D. from Indiana University. His research interests in-clude strategic entrepreneurship, entrepreneurial orien-tation, innovation, and competitive strategy.


THE FOLLOWER’S DILEMMA: INNOVATION AND · PDF fileimitator differentially weighs the...

Documents

Transcript of THE FOLLOWER’S DILEMMA: INNOVATION AND · PDF fileimitator differentially weighs the...