Limits to the diffusion of innovation

Limits to the diffusion ofinnovation

A literature review and integrative modelJason MacVaugh

University of Gloucestershire Business School, Cheltenham, UK, and

Francesco SchiavoneDepartment of Business Studies, University Parthenope, Naples, Italy

AbstractPurpose – The purpose of this article is to integrate existing theoretical explanations for innovationdiffusion across the disciplines of marketing, innovation and sociology research.

Design/methodology/approach – Literature reviews and historical case analysis were used tosupport an integrative model.

Findings – Innovation diffusion is affected by technological, social and learning “conditions” whileoperating in the contextual “domain” of the individual, community or market/industry.

Research limitations/implications – The model is drawn from new product development andmarketing theory. Both fields are dominated by the assumption that users adopt new technology tomaximise their utility. Also, the model does not integrate the overlapping effects of the differentcontexts and domains.

Practical implications – The article provides a sound model for orienting new productdevelopment strategy, since it may reduce the risk of low and slow user adoption of radicalinnovations due, for instance, to their technological, social, and cognitive differences with formerproducts. A second critical managerial implication is that technological, social and learning conditionsclearly have an effect on marketing actions and competitive strategies.

Originality/value – The article provides a literature review of resistance to technology adoptionthrough a multidisciplinary lens.

Keywords Diffusion, Innovation, Social networks, Communication technologies

Paper type Literature review

1. IntroductionOne of the least understood areas of innovation diffusion is the non-adoption of newtechnology (Selwyn, 2003). In some cases individuals or groups eschew thefunctionality of technology, regardless of when it was developed (Bruland, 1995). Inothers, existing technology users do not chose to purchase categorically similar (price,function, availability) newer products when they become available. Further, some whohave used a new technology may later become non-users, given their dissatisfactionwith the experience; also known as discontinuance (Kingsley and Anderson, 1998).

This being said, it must be acknowledged that the majority of research on thistopic to date has assumed that for rational or utility maximising consumerseventually new technology will replace old (Davis, 1989; Rogers, 1962, Venkatesh

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1460-1060.htm

Jason MacVaugh wrote sections 1, 2.4, 3, 4, the model and the appendices and FrancescoSchiavone wrote sections 2.1, 2.2, 2.3 and 5.

Limits to thediffusion ofinnovation

197

European Journal of InnovationManagement

Vol. 13 No. 2, 2010pp. 197-221

q Emerald Group Publishing Limited1460-1060

DOI 10.1108/14601061011040258

et al., 2003). But market history has shown that it does not do so easily,automatically, or even completely. For instance, as early as Gilfillan (1935) hadnoted that in the maritime industry some market segments did not replace sailingships (the old technology) even after the emergence of steam ships (the newtechnology) in the nineteenth century, and diesel in the twentieth century. In fact,producers of the old technology continued commercialisation and acceleratedinnovation in response to the threat of the new technology (Gilfillan called thisphenomenon the “sailing ship effect”).

Accepting that diffusion of innovation is neither uniform (Bauer, 1995;Brandyberry, 2003) nor inevitable, the aim of this paper is to investigate limits toadoption that new technological innovations are likely to find from competition withnon-use of technology and/or notionally older technology. To do so the paper beginswith a cross-disciplinary literature review, exposing the perspectives of researchers inthe fields of marketing, new product development and sociology. This provides atwo-part framework for the examination of innovation diffusion and exposes threelevels of analysis for each framework. Thus it is possible to categorise innovationdiffusion as affected by technological, social and learning “conditions” (Schiavone andMacVaugh, 2009) while operating in the contextual “domain” of the individual,community or market/industry.

The conditions and domains that emerge from the review of innovation diffusionare next contrasted with historical examples of technology non-adoption in context.This evidence provides a lens through which potential drivers of non-adoption becomeclearer. Using Rogers’ (1962)[1] Diffusion of Innovation theory as an organisingframework, an integrative model of factors limiting adoption of new technologicalinnovations is posited to explain the possible effect of interactions between conditionsand domains during the process of new product introduction.

The paper concludes with recommendations for researchers, who will likely findboth the model and the appendices a valuable tool for further research, and forpractitioners, who may look to incorporate predictions of resistance into their productinnovation process.

2. Literature review2.1 Introduction and methodThe theory of adoption and diffusion of innovations (Rogers, 1962) is a useful systemicframework to describe either adoption or non-adoption of new technology. Diffusionoccurs progressively within one market (a system of users) when information andopinions about a new technology are shared among potential users throughcommunication channels. In this way, users acquire a personal knowledge about newtechnology. Knowledge is the first step of Rogers’ five-stages process of adoption. Theother four steps are: persuasion, decision (to adopt or to reject new technology),implementation and confirmation. Accepting this framework, non-adoption can beexplained as the final outcome of an individual process of adoption that failed. Rogersargues that a great number of conditions (e.g. personal limitations of the potential user)and/or external obstacles (e.g. ineffective communication channels) may inhibit thesuccess of the adoption process.

Of course the study of which factors can lead technology adoption to success or tofailure is wide and multidisciplinary. It is also important to note that research

EJIM13,2

198

examining the determinants of successful innovation diffusion is both relativelycommon and, for the most part, internally consistent. For example technology adoptionis a popular theme in marketing and new product development (NPD) literature.Conversely, the research examining non-adoption of technology is relatively limited,but better treatments of the subject usually stem from a Sociological perspective.

Many marketing studies have focus on how new technology is perceived byconsumers, which is usually tested by their behaviours and reactions to technologicalinnovation (Moore and Benbasat, 1991) and how these may change with time andexperience (Kim, 2009). A common interest here is the analysis of user demographics(e.g. Laukkanen et al., 2007; Lee et al., 2005) such as age, gender, education and so on, topredict technology adoption (e.g. Morris and Venkatesh, 2000). Similarly, many NPDstudies attempt to discover which stages in the development process or features of anew product are most critical to achieving market success and wide adoption (e.g.Henard and Szymanski, 2001; Moreau et al., 2001). On the other hand, most sociologicalstudies on this subject analyse how technology adoption is affected by thecharacteristics of society in which potential users are embedded (e.g. Selwyn, 2003;Slowlkowski and Jarratt, 2007). According to these studies, understanding therelationships between users may be more critical than factors relating to the productitself (Brown and Duguid, 1991; Haggman, 2009). As Bruland (1995) highlights,resistance to technology is implicitly a study of the “interaction between thetechnology and its social context”.

To integrate these disparate disciplines the review that follows adopts a multilevelinvestigative approach outlining the context of innovation diffusion. Such an approachalso provides the opportunity to generate as large an explanation of the phenomenonas possible. The three types of conditions, and the three types of domains where theseconditions may occur, were chosen as they reflect the micro, meso and macro levels ofanalysis of economic phenomena (Dopfer et al., 2004). In general, technology adoptionis a multidimensional process in which users’ behaviours are affected by a wide set ofconditions. Learning conditions and the individual domain refer to the micro level ofanalysis, as they are useful to understand behaviours of a single technology adopter.Social conditions and the community domain refer to the meso level, as they show howrelationships between users affect adoption behaviours. Technological conditions andthe market/industry domain refer to the macro level, as they are related to the generalfeatures of an economic system (e.g. a nation) and are result of the sum of more micro(single users) behaviours. Further, this review assumes new technology adoption is amultidimensional process. This process rests on users evaluation of both hard and softfeatures (or conditions) of both the substituted technology and the substitutingtechnology. This process of evaluation takes a shape in multiple contextual domains(fields of action and thought).

2.2 Domains of adoptionNew technology adoption can be said to take place within three domains to due to thethreefold nature of most economic phenomena. The market/industry domain is a(macro) domain of new technology adoption. A second (meso) type of dimension relatesto the set of relationships shaping the social system in which the potential adopters arelocated. Finally, individual (micro) dimension is a third level of analysis likely tosupport the understanding of this process.


199

The critical difference between these domains is the point-of-view undertaken toevaluate and think about new technology adoption. Each point-of-view is particular asit is shaped by a set of specific interests, rules and norms of a given field of action andthought. Each one of these points-of-view thus provides a particular “lens” in order toevaluate positively or negatively a condition (e.g. technological complexity) affectingnew technology adoption. For instance, the diffusion of a new technology within aneconomic system in order to satisfy a specific function (e.g. the use of PC to watchentertainment programmes through web-broadcasting in substitution of television andtraditional broadcasting) is considered, evaluated and accepted (or rejected) in threedomains at the same time:

(1) In the domain of the industry/market, in which the benefits and costs of changeare evaluated according to their impact across a large economic system. Forexample, how might web broadcasting affect the business of traditional TVchannels or behaviours of TV viewers? What technological infrastructurescould a country provide to increase utility for users and firms adopting webbroadcasting?

(2) In the domain of community of users, in which the benefits and costs of changeare evaluated according to their impact on social relationships betweencommunity members. For example, how web-broadcasting may change powerrelationships within local politics.

(3) In the domain of the single user, in which the benefits and costs of change areevaluated according to their impact on the personal utility of a single user. Forexample, what skills and capabilities should the single user acquire in order toadopt and utilise web broadcasting?

Such a process of diffusion thus involves three different fields of thinking andevaluation. Of course, a unique condition may produce divergent and/or similarevaluations across domains. For instance, network externalities are a typicaltechnological condition that significantly affects the adoption and diffusion of aninnovation at both individual and industrial level. This construct refers to the utility anindividual adopter of an innovation achieves by the increase in total number ofadopters in the technological market (Shapiro and Varian, 1999). Classic examples ofhow network externalities facilitate the diffusion of a new technology include thetelephone, fax or internet. Similarly, technological conditions are critical in the so-callednetwork industries (e.g. broadcasting, personal computing and air transportation). Anetwork industry is a market with the following characteristics: complementarity,compatibility and standards; consumption externalities; switching costs and lock-in;and significant economies of scales in production (Shy, 2001). The diffusion of aninnovation is only possible in this type of industry if specific complementarytechnological infrastructures (e.g. televisions, personal computers, or airports) areavailable and work (Shy, 2001).

Similarly, several studies explore the importance of communities of practice forindividual learning and innovation adoption (Brown and Duguid, 1991; Wenger, 1998).A community of practice “defines itself along three dimensions:

(1) What it is about — its joint enterprise as understood and continuallyrenegotiated by its members;

EJIM13,2

200

(2) How it functions — the relationships of mutual engagement that bind memberstogether into a social entity;

(3) What capability it has produced — the shared repertoire of communalresources (routines, sensibilities, artefacts, vocabulary, styles, etc.) thatmembers have developed over time”.

Individuals who decide whether or not to adopt a given technology tend to act andexchange information within one or more social communities to which they belong.The adoption of a technological innovation is thus dependent on an individualsense-making process that a potential adopter undertakes every time he recognisesthat an innovation may satisfy their needs and be socially accepted and awarded bytheir community.

At individual and community domains, users knowledge has an importantinfluence on the market technological change as it mediates individual choice for theadoption of an innovation. Shapiro and Varian in their work on information economy(Shapiro and Varian, 1999) use the concept of switching costs to stress how the extentof knowledge and capabilities of individuals using an existing technology can hampertheir adoption of a new technology aimed at substituting the former one due to a“lock-in” constraint. Switching costs are generally defined as the costs that a consumerattracts when they decide to switch to a competitor’s product or service. They generatea lock-in effect, arising “whenever users invest in multiple complementary and durableassets specific to a particular information technology system” (Shapiro and Varian,1999, p. 12). Conversely, Moreau et al. (2001) highlight that the product class knowledgeheld by existing users can provide a distinct advantage in understanding the value ofthe innovativeness of a new product. Therefore, it is no surprise that recentinvestigations of high technology product introduction suggest the need for producerdriven training to be provided (Hanninen and Sandberg, 2006).

2.3 Conditions of adoptionTechnological conditions help explain technical and market features of thesubstituting technology and the substituted product. New products entering into amarket are rarely completely new, more often their design arises from other substitutorand complementor technologies or market products (Brandenburger and Nalebuff,1995). The availability of complementary technologies positively affects the adoptionof new substituting technology (Teece, 1986; Gandal et al., 2000). For instance, the rateof adoption of USB pen-drives (technological device substituting hard-diskettes) wasstrongly dependent on the prior diffusion of USB ports in the personal computermarket. Similarly, the diffusion of operating systems has historically been correlatedwith the amount of software available for it to run. For example, most releases ofopen-source operating systems, such as Ubuntu, comprise compatible free softwarepackages as well.

Given the complex nature of high technology products, most of their developermarkets consist of industrial networks. For users, this results in a need to access agreater number of technologies in order to utilise a single product. When thisprocess brings about the presence of an “industry standard”, there is usually areciprocal reduction in the adoption of radical technological innovations within thatmarket. Indeed, an industry standard links together a network of additional


201

complementary technologies (e.g. CD players, compact discs, and software). Thismakes it harder for users of an existing technology (and its technological network)to adopt a newer and completely different product satisfying the same needs, as theutilisation of the innovation requires complementary technologies not yetwidespread in the market.

Furthermore, the technological complexity of an existing and widely adoptedproduct reduces its retire-ability as well. If a complex product is an artefact bridgingtogether more levels of technologies (each one with specific design settings) (Murmannand Frenken, 2006), then users may find it both risky and expensive to shift to adifferent technology made by technological subsystems utilising different components.For instance, car drivers often prefer to buy cars with major maintenance supportsupply chains in their own country, as this makes it cheaper and faster to maintain thecar as it ages. This brings about certain user expectations of the success of anemerging technology; thus its network of complementary products will affect its rate ofadoption (Gandal, 2002).

Social conditions explain the cultural and relational specificities (e.g. norms, values,hierarchies) widely shared within the groups or communities to which users belong.For instance, the status that users acquire within their own social group by using agiven technology influences their propensity to change it for newer products. From thisperspective, if a member of a community of vintage cars installs a modern CD player,other members probably would “disapprove” of this change and consider them, andtheir car, less worthy of being a community member.

This risk probably might be lower if the adopter was a community opinionleader. This kind of individual plays a critical role in the diffusion of innovations.Opinion leaders are individuals who frequently influence others’ orientations towardadopting an innovation (Rogers, 1962). If the use of a technological innovation isnegatively accepted or misunderstood within a community, the rate of its adoptionis likely to be slowed too. Firms usually ask for opinion leaders support in order toprevent this risk. A positive appreciation of opinion leaders (if they exist) is criticalfor expanding the social adoption of technological innovation within theircommunity or market.

However, diffusion of innovation can also be considered a “bandwagon” processdeveloping within a social network and relying on reciprocal contagion between its“peer” nodes (e.g. Abrahamson and Rosenkopf, 1997). Social contagion is the processby which a person catches an idea or behaviour from another person (Burt and Janicik,1996). It is a specific feature of networks and is commonly operationalised throughcohesion and structural equivalence, two typical network measures considered as thedriving mechanisms of contagion. Medical innovation is a well-known example of howsocial contagion within a community of users can mediate the dynamics of newproduct adoption (Coleman et al., 1966). This being said, studies have also shown howboth contagion and personal preferences of doctors are equally critical to orientadoption of innovations in medical communities (Burt, 1987).

Learning conditions are individual characteristics of a single user. These conditionsare likely to affect the acquisition of new competencies and capabilities necessary touse a new technology. However, learning is also a multi-step social process throughwhich an individual (or an organisation) acquires codified information and/or tacitknowledge from its external environment, internalises this new bundle of knowing,

EJIM13,2

202

and utilises it in order to innovate its actions. Socialisation is a critical phase of thisprocess (March, 1991; Nonaka, 1994) and it depends on the extent of social contagion ofthe user’s community. However, adopters must be able to absorb external knowledgeand apply it in order to utilise new technology. Thus, the extent of the single usersabsorptive capacity (Cohen and Levinthal, 1990) can positively affect their learning ofhow to use a new technology and make it less difficult to retire an existing one.

Furthermore, the extent of the switching costs that a potential adopter of a newtechnology has to afford in order to learn how to utilise the new one depends onhow much time and effort this individual spent learning how to use the oldtechnology and its features. For instance, a typical advantage of first movers innetwork industries is their capability to establish a dominant design, which canquickly enter a market, making it harder for competitors to gain market share withalternative products afterwards. A dominant design is a product widely adoptedwithin the corresponding industry and its emergence apparently changes the natureof the market competition (Abernathy and Utterback, 1978). The main implication ofthis is the first mover product becomes the Industry Standard its competitors areforced to follow. A similar situation characterises the software industry whereMicrosoft with its suite (Microsoft Office) was the first mover (1989) in the marketsegment of office suites, and today it is still the leading design standard betweenthe various office suites. Indeed, Microsoft launched the programme Word 1.0 forMacintosh in 1984 (first year of commercialisation of this computer) before packingit with other applications (as Excel or PowerPoint) into the first Office suite fiveyears later.

2.4 Exposing patterns of non-adoptionWith grounding in the conditions preceding technology adoption, and with a soundunderstanding of the contextual domains in which technology is adopted, the literaturereview thus far provides solid foundations on which to analyse patterns of technologyuse through a new lens of non-adoption. In this regard, it seems only logical to beginwith by far the most often noted driver of technology adoption, technological utility.This being said, it is almost counterintuitive that products would be developed orreleased to market without superseding the utility of the existing product ortechnology, but it is not without precedent in recent history. Of course even the mostcommitted technophile would agree that utility is in the eye of the beholder. Forexample, products developed for our western desire for disposability and speed, suchas the microwave-meal, often proven to be of a lower quality than their home orrestaurant produced equivalent. Here, utility might be measured positively when timeis short, with taste becoming a less important factor. In this instance, most consumerswould and do revert to the older technology (oven) or process (“from-scratch”) whenmeans or time allow.

Instances of quantitatively poor utility in high-technology development are, bythe nature of the process, rare. So given its “green” technological utility, it isreasonable to wonder why advancements in electric engine and battery design havenot resulted in the demise of petrol engine automobiles? It is not for lack of “new”technology, of which there is plethora. The significant problem electric cars face isthat users construe performance based on the existing effectiveness of the currentlyavailable product (Moore and Benbasat, 1991). From a user perspective even very


203

poor performing petrol engine cars cost less, drive further, and are cheaper tomaintain than the best electric cars. So while it can be assumed that there maycome a time when all of the various technologies needed to make electric cars cheapand with respectable range will come, it is important to highlight this is notdependant on developments in electric engine and battery technology alone. Thusolder technologies are usually not replaced by newer when the technologicalcomplexity of the product category tends to focus evaluations on overalleffectiveness (Moreau et al., 2001) rather than the utility of newer features.

Another interesting case of simple technological utility not driving consumptionoccurs in the personal computer (PC) market. Scanning popular technology periodicals(for example Smith, 2007) and national newspapers, Apple all-in-one personalcomputers are often rated as meeting user needs in key areas more reliably, and withtechnology ready to use “out of the box”, than comparable Windows based PCs.Following either the utility maximising or technological complexity perspectives ontechnology adoption, Apple could, arguably, become the dominant player in the homecomputing market. But, one important factor accounting for the 90 percent share heldby Wintel (Windows operating system/Intel processing chip based) PCs is theoverwhelming advantage in the availability of complementary technology such asbusiness related software, gaming software, peripherals, replacement parts andsupport available from local technicians. Thus, for non-expert users the utility of PCtechnology may be limited, but it is greatly extended by the complementarity of itsrelated industrial network (Shy, 2001). In this case new technology fails to replace oldertechnology when technological complementarity creates higher total utility for existingusers than would be gained by adopting the new technology.

Utility, regardless of its nature or source, has never been a complete explanation forthe behaviour of humans. In many circumstances it is not utility but the context thatdetermines use of new technology. Selwyn (2003) argues that by far the most extensiveresearch explaining technology non-adoption does so through discourses of userdeficiency. In such instances it is the social context; for example deficient materialwealth (Krieg, 1995) in poor communities, enforced social demarcation (Chatman, 1996)in the workplace, or authority driven exclusion (Taylor et al., 2003) in the case ofsensitive or dangerous technologies, which explains non-adoption far better thanutility of the technology in questions.

Undoubtedly though, there are many contexts in which technology is a welcomedaddition to social life, for example regardless of a community’s wealth, most peoplehave an appreciation and some form of access to recorded music. Nevertheless, forthe wealthy audiophile, one who has developed enduring involvement (Richins andBloch, 1986), music appreciation also extends to the technology used to recreate themusic. The result within the related “hi-fi” market/industry has been a drive toreduce distortion in recording and increase the ability to hear the full range of tonesemitted. There are, of course, limiting factors in this reproduction, of which the mostimportant is the change in the characteristics of sound as it is amplified or captured.Many audiophiles share a strong conviction that the amplification and captureprocess should utilise “tube” technology, so named for vacuum tube circuits. Earlyresearch suggested that the preference for such technology is not limited to simplequalitative appreciation for music from tube systems, but that they also havespecific and measurable effects on the characteristics of the recorded sounds

EJIM13,2

204

(Bussey and Haigler, 1981). In the last 15 years the capability of digital equipmentto model and replicate “tube sound” has dramatically increased, effectively wipingout the measurable differences between tube and transistor amplification or capture.So while scientist might reasonably accuse audiophiles of having generated someform of groupthink (Janis, 1972), it must also be acknowledged that users developtheir own patterns for making sense of technology (Seligman, 2006) and that it isthis process that forms the grounding for evaluations, not bench science.Accordingly, users of tube technology continue to hold negative stereotypes aboutdigital technologies, many of which stem from weaknesses highlighted during thetechnology’s introduction over 30 years ago. So here it can be argued that,regardless of simple social context, older technologies survive when the socialorientations towards the newer technology are negative.

In some cases, though, it is not an excluding social context or a negative socialorientation towards technology that results in non-adoption, but rather its social uses.For example, information communication technology (ICT) has supplanted so manyother existing technologies, both at home and in the workplace, that it is almost moot tocodify the extent of its adoption. Furthermore, in some high technology “corridors”such as Silicone Valley (California, USA) ICT has also become a necessary feature ofday-to-day life. However, despite the oft-lauded advantages of such technology, manycommunities in other high technology corridors, such as Bangalore in India, eschewICT’s social uses, such as dating, food delivery and social networking. The interestingfactors being that the knowledge workers of Bangalore are neither cognitively normaterially deficient with regards to ICT. As many of them have trained for many yearsto work in the ICT industry, it can also be assumed that they do not hold negativestereotypes of the technology. Here, Bruland (1995) argues that non-adoption shouldrather be seen as a positive part of the social selection process. So, to study the strengthof the contagion of a new technology it is also necessary to understand the localcultural characteristics of the intended user groups. In this case older technologiessurvive when the new technology does not create a strong enough social contagion todisplace the community of non-users.

Of course while social factors have in the past resulted in a desire to own a newtechnology, one key axiom is that the owner will require some level of new learning toenable use. Thus the user, user community or technology provider must negotiate thebarrier of knowledge capacity. The ability of a human or groups of humans to learn is,of course, limited (Cohen and Levinthal, 1990). Access to learning may be limited in agiven community (Miller, 1994), and in some instances is restricted by the failure of thetechnology-developing organisation itself (Hanninen and Sandberg, 2006).

While in recent years governments and educators have sought to increase thelearning capacity of potential ICT users, Bower and Christensen (1995) argue thatmany successful innovating organisations fail to recognise the effect that really newtechnologies will have on existing ICT users. Introducing a radically new product canbe made much more difficult if the product requires extensive marketing for consumersto understand that the new product is in the same product class as the old (Brucks,1985). This being said, some innovations never fully disperse into their intendedmarket and may never displace established devices. Good examples of thisphenomenon are digital input technologies such as voice or handwriting recognition.While there are many good technical rationales supporting the development of such


205

technology (allowing support for disabled users, speed of use, simulation of naturalhandwriting etc.) traditional typing surfaces or telephone keypads remain by far themost dispersed and adopted in the ICT market. One straightforward explanation forthe relatively low uptake of these newer technologies is that being able to use akeyboard or number pad does not in any way prepare users for “talk-type” microphonesystems or “stylus” based handwriting recognition devices. Most such devices requirethe user to learn the correct input technique. Hanninen and Sandberg (2006) go so far asto suggest that high technology manufacturers should set out a “roadmap” to ensureend users learn enough to be able to use the technology in new high-technologyproducts. Nonetheless, such an educational framework is of little interest to the passivetechnology user (Alba and Hutchinson, 1987), and to some extent this explains thecontinued popularity of a number of notionally out of date technologies. In thisinstance older technologies survive when existing learning capabilities do notsignificantly assist in use of the new technology.

Nonetheless, many individuals and organisations are more than willing to developthe capability to use a product if the perceived utility (Venkatesh et al., 2003) is highenough. One example of a product providing irresistible utility is a bespoke corporateICT system. Well-known ICT systems, such as the SABRE ticket booking system firstdeveloped for American Airlines by IBM in the 1950s, have created a distinctcompetitive advantage. In the SABRE case, the advantage generated by an innovativeICT system, and the relatively late uptake of such technology by others, is oftenattributed to the collapse of several competitors in the market (such as BraniffInternational Airways). However, such systems represent one of the largest singleinvestments that an organisation can make. These costs arise not only from thedevelopment of the software, but also from the installation of appropriate hardware,group specific configurations, user training programs, and ongoing systemmaintenance. Such investments in learning, management time, and capitalexpenditure are not made up lightly. So it is quite common for such systems toremain in use when newer technology has long surpassed the older technology’scapabilities. Therefore, one final explanation for this continued use of an oldertechnology is the extent to which a new technology requires switching costs (Shapiroand Varian, 1999) higher then the perceived utility gained by its use. In some casesolder technologies survive when the learning costs that would be incurred byswitching from old to new technology are prohibitively high.

3. An integrative model of non-adoptionOur research indicates that in any system of innovation diffusion there are threecommon players: the (potential) individual user, the community of users (ofcategorically similar technology) and the innovating industry or market. The playersinfluence and are in turn influenced by three systemic conditions: technological, socialand learning. Conditions have an effect at a number of levels, each of which we haveidentified as a variable associated with adoption/non-adoption. In any systemicinstance of product introduction where a player is influenced by a condition resultingin negative feedback (Sterman, 2001) the result may be non-adoption. In a systemicinstance of product introduction where multiple players and condition interactionsresult in negative feedback, the result is almost certainly non-adoption.

EJIM13,2

206

Figure 1 follows the pattern of analysis from the literature review. Starting on thehorizontal axis we highlight that technology non-adoption usually takes place withinat least one of three domains: that of the individual, a community of individuals orwithin a marketplace/industry. On the corresponding vertical axis we show hownon-adoption may be accounted for by consideration of one or more conditions:Technological, Social structure or Learning. Each of these conditions has at least three,increasingly complex, levels of influence. For example it is much simpler to identifywhen a technology is of a lower utility to a single customer than it is to know when anentire community favours a product because of the complementarity generated byindustry standards. Thus, reading the model from top to bottom, increasingly complexreasons for non-adoption are exposed. Similarly, reading from left to right the contextmoves further away from individual domains towards those that are bound up in theindustry or market as a whole.

Advice on reading the model: reading from the upper left hand text box, thenchoosing one of the contextual factors boxes, and finally moving across to a chosendomain, it is also possible to read the model as several short summaries of ourliterature review. For example, a new product development team might wish tounderstand why a community of users did not accept its newest product. Given datafrom a marking survey they might find out that users had not been able to find outmuch about the product from a popular online forum for related technologies. Readingfrom one line of the model they could surmise:

New technology fails to replace older (or no use of) technology when the capability of usershas not created a community of expertise, as can be supported by the research of Aggarwalet al. (1998) and Maryse and Eelko (2008).

4. LimitationsWhile we believe the model provides a much-needed integration of the highly dispersedknowledge on the subject of technology adoption, we must also acknowledge somelimitations of our approach. First, much of the work on which we based the model isdrawn from NPD and Marketing theory. Both fields are dominated by the assumptionthat users adopt new technology to maximise their utility. Conversely, the Sociologyliterature argues that consumers may adopt a new technology, for instance, byfollowing a temporary fashion. In this case the user is not maximising their utility, butrather maximising their social orientation. The resulting conflict of assumptions,means that the model is context dependent rather than generally predictive.

Second, the model does not integrate the overlapping effects of the different contextsand domains in which almost all new technology operates. Thus, while each centralbox explains one important reason for non-adoption it does not explain how this mayor may not be related to other reasons. In the tradition of Innovation research this isuseful as it provides a sequential system for the consideration of strategic choices, butSociologists may find such explanations overly simplistic without discussion of thesystem as a whole.

Finally, we must highlight that our model is based on a comparison of existingtheory with historic data on technology non-adoption. We have no direct evidence thatthe factors discussed in our case examples caused the non-adoption, nor that theoriginating authors of the research we apply would accept our interpretation of their


207

Figure 1.An integrative model offactors limiting innovationadoption

EJIM13,2

208

results. The literature review and model is exploratory, illustrative, theoretical innature, and requires quantitative study to verify its claims.

5. Conclusions and recommendationsThis paper proposes an integrative model aimed at explaining limits to adoption ofinnovation and non-adoption of new technologies. The model is built on theassumption that individual decision to adopt or reject new technology is affected byseveral variables. In brief, they can be summarised as:

. the extent to which new technology meets the significant technological, social,and/or learning conditions encouraging its adoption i.e. it is easy to adopt; and

. the extent to which new technology is considered useful in the individualdomain, community domain and/or industry/market domain; i.e. it is useful.

The main conclusion emerging from our literature review (see Tables I-VI for indexesof the studies reviewed for our model, shown in chronological order) and theoreticalmodel is that firms launching product innovations should consider a broad range ofvariables in order to maximise their adoption within market. Some of these variablesrefer directly to the new technology (e.g. price, perceived utility, diffusion rate,technological infrastructures and so on). Other variables affecting adoption are instead“external” to new technology as they related to the characteristics of old technologythat innovation should replace. When users evaluate these characteristics as moresuitable (for instance, in terms of social orientation, utility, technologicalcomplementary and so on) than characteristics of recent innovations, the moreproduct adoption is restrained. This reasoning also highlights the need for academicsto study technology adoption through a multidisciplinary lens (a useful starting pointis shown in Tables VII-VIII, which show meta-analyses and overview texts in the field,in chronological order). This would increase their understanding of existingphenomena and the likelihood of successful prediction of future innovation success.

Two main managerial implications can be drawn. First, this can be a fruitful modelfor orienting new product development strategy, since it may reduce the risk of lowand slow user adoption of radical innovations due, for instance, to their technological,social, and cognitive differences with former products. A second critical managerialimplication arises for strategy and marketing boards of high-tech firms. Technological,social and learning conditions clearly have an effect on marketing actions andcompetitive strategies of this type of organisation. For instance, if some firms perceiveusers switching costs (or community resistance) for the adoption of a new technologyare too high, they should not risk engaging in direct competition with innovationpioneers. Conversely, they could exploit such conditions and seek new opportunities,for instance, by launching products aimed at revitalising a mature technology.Similarly, they could arrange marketing actions (e.g. media campaigns) in order toreinforce the social adoption and utilisation of the old technology within its usercommunity.

These implications also provide the basis for further empirical research. Theauthors plan to conduct surveys in order to investigate to what extent conditions anddomains matter in retiring an old product and accepting a new one. Furthermore, werecommend that others examine the potential interdependencies and regularitiesbetween the model variables and domains.


209

Abernathy

and

Utterback

(1978)

Maidiqu

eandZirger

(1984)

Richins

andBloch

(1986)

AlbaandHutchinson

(1987)

Burt(1987)

Dom

ain

Market/indu

stry

Com

mun

ity/market/

indu

stry

Individu

alIndividu

al/m

arket/

indu

stry

Individu

al/com

mun

ity

Conditions

Techn

ology

Techn

ology

Social

structure

Learning

Social

structure

Objectives

Und

erstandpatterns

ofindu

strial

innovation

Examinedriversof

productinnovation

success

Und

erstandtemporal

contextof

product

involvem

ent

Examinationof

consum

erkn

owledg

econstructs

Examinedriversof

social

contagion

Method

Empirical

Empirical

Empirical

Theoretical

Empirical

Results

Innovation

withinan

establishedindu

stry

isoftenlim

ited

toincrem

ental

improvem

ents

ofboth

products

andprocesses.

Dom

inantdesign

affects

technology

substitution

Listof

technology

specificsuccess

determ

inants.H

ighlights

specialchalleng

esin

the

developm

entof

“high-

tech’p

rodu

cts

Consumer

with

End

uringInvolvem

ent

exhibitstable

involvem

entbehaviours

even

whennot

purchasing

orusingthe

product

Consumer

know

ledg

eis

complex,twomain

components;familiarity

andexpertise.Deficiency

ineither

hasan

adverse

affect

onchoice

toconsum

e

Twomajor

factorsin

contagionare

equivalenceand

cohesion.T

hefirstis

strong

erfor

professionalswho

are

likelyto

adoptan

innovation

when

appropriateto

commun

itycontext

Specificmention

oftechnology

non-

use?

No

No

No

No

Yes

Table I.

EJIM13,2

210

StuartandAbetti(1987)

Zeithmal

(1988)

Davis(1989)

Cohen

andLevinthal

(1990)

BrownandDug

uid

(1991)

Dom

ain

Market/indu

stry

Individu

alIndividu

alIndividu

al/m

arket/

indu

stry

Individu

al/com

mun

ity

Conditions

Techn

ology/social

structure

Techn

ology

Techn

ology

Learning

Learning/social

structure

Objectives

Predict

technology

start-

upcompany

success

Und

erstandconsum

ers

perceptionsof

quality

andvaluein

relation

toprice

Createvalid

measurementscalefor

user

adoption

ofIT

Und

erstandaffect

ofabsorptive

capacity

onlearning

andinnovation

Und

erstandhow

commun

itiesaffect

innovation

andlearning

inworkp

lace

Method

Empirical

Theoretical

Empirical

Empirical

Theoretical

Results

Entrepreneur

involvem

entand

pursuing

new

ventures

that

match

existing

operationalexperience

arethekeydriversof

success

Consumer

perceptionsof

aproduct’s

qualityand

valuearemorecomplex

than

indu

stry

standard

measuresandpu

rchase

price

Useradoption/utilitycan

beaccurately

measured

byperceivedusefulness

andperceivedease

ofuse

Capacityforan

individu

alor

anorganisation

tolearnis

limited.A

bsorptive

capacity

isin

part

determ

ined

byexisting

know

ledg

e

Capacityfora

commun

ityto

affect

social

orientations

and

learning

practicesof

its

individu

almem

bers

towards

new

andold

technologies

Specificmention

oftechnology

non-

use?

No

No

No

No

No

Table II.


211

Ellenet

al.(1991)

Moore

andBenbasat

(1991)

Fornell(1992)

Calantone

etal.

(1993)

Miller

(1994)

Bruland

(1995)

Dom

ain

Individu

alIndividu

alMarket/indu

stry

Market/indu

stry

Individu

al/

commun

ity

Individu

al/

commun

ity

Conditions

Learning

Techn

ology

Learning/technology

Techn

ology/social

structure

Learning

Social

structure

Objectives

Und

erstandroleof

ability

and

satisfaction

inresistance

totechnology

use

Develop

methodto

measure

perceptions

oftechnology

before

adoption

Createandexplain

thefactorsinvolved

inacustom

ersatisfaction

barometer

Exp

lain

keystepsin

new

productlaun

chandfollow

onactivities

Exp

lore

gend

erdifferencesin

training

access

and

uptake

Examinepatterns

ofresistance

tonew

technology

Method

Empirical

Empirical

Empirical

Theoretical

Empirical

Empirical

Results

Aperson’sperceived

ability

tousea

productaffectstheir

response.T

helevel

ofsatisfaction

experiencedwithan

existing

product

increasesresistance

toadopting

analternativeproduct

New

technology

utility

canbe

evaluatedby

locating

user

perceptionsviaa

seven-category

scale.Non-adopters

have

sign

ificantly

differentscale

profi

les

Customer

satisfaction

isim

portantin

heterogeneous

markets.S

ection

onim

portance

oforganisational

responsesto

switchingbarriers

such

aslearning

and

cost

Produ

ctlaun

chand

commercialization

requ

ires

themost

time,money

and

attention.

Itisa

concurrent

activity

that

sign

ificantly

affectsup

take

bychannels,retailers

andconsum

ers

Wom

enandlower

levelworkers

are

less

likelyto

beofferedaccess

totraining

inthe

workp

lace.S

uch

training

isim

portant

fordeveloping

equity/equ

alityin

theworkp

lace

Manyoverlapp

ing

layers

(personal,

relig

ious

andlegal)

ofsocial

resistance

may

preventanew

technology

from

beingaccepted

byindividu

alsand

commun

ities

Specific

mention

oftechnology

non-

use?

Yes

Yes

No

No

No

Yes

Table III.

EJIM13,2

212

Krieg

(1995)

Rogers(1962)

Chatm

an(1996)

Agg

arwal

etal.

(1998)

Kingsleyand

And

erson(1998)

ShapiroandVarian

(1999)

Dom

ain

Com

mun

ity/market/

indu

stry

Individu

al/

commun

ity/market/

indu

stry

Individu

al/

commun

ity

Com

mun

ity

Individu

alMarket/indu

stry

Conditions

Social

structure

Techn

ology/social

structure/learning

Social

structure

Learning

Social

structure

Techn

ology/learning

Objectives

Exp

lore

access

totechnology

inpoorer

urbancentres

Exp

lain

diffusionof

innovations

Exp

lore

“insider”/

”outsider”

dynamics

incommun

ities

Und

erstand

adoption

of“really

new

products”

Und

erstandnon-use

ofinternet

bythose

who

have

triedit

Exp

lain

econom

icfactorsregu

lating

theinform

ation

econom

y

Method

Theoretical

Empirical

Empirical

Theoretical

Theoretical

Theoretical

Results

Manyurbancentres

arepoor,w

ith

inequitablematerial

access

totechnologies.

Conversely

inform

ation

technology

hasthe

potentialto

alleviate

thisposition

Bookanalyses

innovation

diffusion.

Includ

esim

portant

chapteron

diffusion

netw

orks.H

ere

opinionleadersand

commun

ities

strong

lyinfluence

technology

adoption

Inform

ationaccess

isboun

dedin

social

context.Insiders

determ

ine

“app

ropriate’

know

ledg

e/technology

.Outsiders

rely

onstrategies

todeal

withinform

ation

poverty

Really

new

products

requ

ireconsum

erlearning

.Without

“surrogate

buyers’

thislearning

isless

likely

Somehave

never

used

theinternet,

andothers

have

nomaterialaccess,b

utthereisagrow

ing

numberwho

have

discontinu

eduseand

donotintend

toreturn.T

hisiscalled

internet

“chu

rn”

Bookdiscussesof

inform

ation

econom

icsand

strategy

.Includes

andim

portant

discussion

for

managem

entin

and

around

“techn

ology

lock-in

”markets/

indu

stries

Specific

mention

oftechnology

non-

use?

Yes

Yes

No

No

Yes

No

Table IV.


213

Song

andMontoya-

Weiss

(1998)

Wenger(1998)

Klin

g(1999)

Easingw

oodand

Koustelos

(2000)

Morrisand

Venkatesh

(2000)

Moreauet

al.(2001)

Dom

ain

Market/indu

stry

Com

mun

ity

Individu

al/m

arket/

indu

stry

Market/indu

stry

Individu

alIndividu

al

Conditions

Techn

ology

Social

structure

Social

structure

Techn

ology/social

structure/learning

Social

structure

Techn

ological/

learning

Objectives

Und

erstand

developm

entof

new

VsIncrem

ental

technology

Exp

lore

importance

ofcommun

itiesof

practice

Relationship

betw

een“new

internet’and

ordinary

users

Und

erstandpost-

developm

entnew

productmarketing

Und

erstand

influenceof

ageon

technology

adoption

Und

erstand

relationship

betw

een

consum

erperception

andkn

owledg

ebase

Method

Empirical

Theoretical

Theoretical

Theoretical

Empirical

Empirical

Results

Successdriversare

differentdepend

ing

onproduct

innovativeness.

Really

new

products

requ

iremoretime,

resourcesandeffort.

Process

successis

also

more

unpredictable

The

commun

ityof

practice

isa

netw

orkedsystem

regu

lating

group

behaviour,

influences

individu

alsandtheir

decisions,and

legitimises

relationshipsand

authority

For

ordinary

usersto

benefitfrom

investments

in“new

internet’

technologies

materialandaccess

barriers

need

tobe

considered

and

removed

Preparation

begins

withnetw

ork

build

ingandsupp

ort

forlearning

.Targeting

different

elem

ents

ofthe

social

structureis

key.

Poorexecution

ofismoresign

ificant

than

thetechnology

Age

isasign

ificant

antecedent

factor

indriversof

technology

adoption.

Regardlessof

age,

perception

oftechnology

isthe

strong

estdriver

Knowledg

ein

the

base

domainmay

have

anegative

influenceon

preference

for

discontinu

ous

innovations

Specific

mention

oftechnology

non-

use?

No

No

Yes

No

No

Yes

Table V.

EJIM13,2

214

Shy(2001)

Gandal(2002)

Halland

Khan(2003)

Tayloret

al.(2003)

Hanninenand

Sand

berg

(2006)

MaryseandEelko

(2008)

Dom

ain

Market/indu

stry

Individu

al/

commun

ity/market/

indu

stry

Market/indu

stry

Market/indu

stry

Market/indu

stry

Com

mun

ity

Conditions

Techn

ological

Techn

ological

Techn

ological

Techn

ological

Learning

Learning

Objectives

Analyse

and

Und

erstand

NetworkIndu

stries

Codifynetw

ork

theories

toinform

policydecision

making

Und

erstandfactors

affectingnew

technology

adoption

Und

erstandeffects

ofgovernmental

decisionson

innovations

Und

erstandfirm

’sability

toevaluate

andgu

ideuser

learning

Und

erstandeffect

ofcurrente-commerce

adoption

onfuture

intentions

Method

Empirical

Theoretical

Theoretical

Empirical

Theoretical

Empirical

Results

International

standardization

affectspositively

new

technology

adoption.Ifitis

missing

then

old

technology

may

survivelong

er

Consumer

expectations

and

marketcompetition

redu

cethenu

mberof

netw

ork

technologies

over

time.There

isa

necessarytradeoff

betw

eenprivateand

social

driversof

adoption

Regulatory

environm

entand

governmental

institutions

have

apowerfuleffect

ontechnology

adoption,

oftenviatheability

ofagovernmentto

“sponsor”a

technology

Governm

ent

regu

lation

appears

tobe

agreater

stim

ulus

toinventive

activity

than

government-

sponsoredresearch.

The

anticipation

ofregu

lation

also

spurs

inventiveactivity

Recom

mends

firm

splan

involvem

ent

withconsum

erlearning

byuseof

a“roadm

ap”

The

less

know

ledg

ecompanies

have

aboute-commerce,

theless

likelythey

areintend

toadopte-

commerce

Specific

mention

oftechnology

non-

use?

No

No

No

No

No

No

Table VI.


215

Soud

erandSh

erman

(1993)

Montoya-W

eiss

andCalantone

(1994)

BrownandEisenhardt(1995)

Bauer

(1995)

Considers

domains

Com

mun

ity/market/indu

stry

Com

mun

ity/market/indu

stry

Individu

al/com

mun

ity/

market/technology

Individu

al/com

mun

ity/

market/technology

Considers

cond

itions

Techn

ology/social

structure

Techn

ology/social

structure

Techn

ology/social

structure/

learning

Techn

ology/social

structure/

learning

Objectives

Provide

overview

ofnew

products

developm

entprocess

managem

ent

Meta-analysisof

new

product

developm

entsuccessdrivers

Modelproductdevelopm

ent

successfactorsfrom

multiple

actorperspectives

Exp

lore

resistance

tonew

technology

Method

Theoretical

Empirical

Theoretical

Empirical

Results

(Book)

Providesdetailedlook

atstepsin

new

product

developm

entprocessfrom

the

perspectiveof

theinnovating

organisation.H

ighlightsbest

practicesandcases

Highlightsthat

research

todate

hasprovided

anextensivebu

tnotyet

conv

ergent

setof

success

factorsfornew

product

developm

ent

Literatureon

thetopiccanbe

categorisedas

rational

planning

,com

mun

ication

linkedor

disciplin

edproblem

solving.

Eachexpose

behaviourof

actors

andtheir

affect

ondevelopm

ent

outcom

es

(Book)

provides

insigh

ton

perspectives

(oftheInnovator,

Resistant

andObserver)and

objects(others,selfand

system

)as

determ

inants

ofnew

technology

resistance

Specificmention

oftechnology

non-use?

No

No

No

Yes

Table VII.

EJIM13,2

216

HenardandSzym

anski(2001)

Selwyn

(2003)

Venkatesh

etal.(2003)

Galende

(2006)

Considers

domains

Com

mun

ity/market/indu

stry

Individu

al/com

mun

ity

Individu

alMarket/indu

stry

Considers

cond

itions

Techn

ology/social

structure

Techn

ology/social

structure/

learning

Techn

ology/social

structure/

learning

Techn

ology/social

structure

Objectives

Meta-analysisof

new

product

developm

entsuccessdrivers

Develop

aricher

pictureof

ITnon-use

Unified

modelof

I.Tadoption

Examines

literatureto

high

light

meta-approaches

toinnovation

theory

build

ing

Method

Empirical

Theoretical

Empirical

Theoretical

Results

Highlightsthat

research

todate

hasprovided

anextensivebu

tnotyet

conv

ergent

setof

success

factorsfornew

product

developm

ent.Further,the

papercastsdoub

tson

determ

inantsmeasure

inother

new

productdevelopm

ent

stud

ies

ITuseistheresultof

goodness

offitbetw

eenthe

technology

andtheperson’s

life.Thisin

turn

isinfluenced

bycomplex

structures

and

agentsboth

inthepastandthe

present

Modelof

interrelationships

betw

eendeterm

inants

ofan

individu

al’sI.T

use:

performance

andeffort

expectancy,socialinfluence,

cond

itions,d

emograph

icsand

experience

Eachof

thefive

meta-

approaches

(Ind

ustrial

organisation,T

CE,A

gency,

RBV,E

volutionarytheory)are

limited,b

utEvolutionary

theory

hasthehigh

est

explanatorycapability

Specificmention

oftechnology

non-use?

No

Yes

No

No

Table VIII.


217

Note

1. Theory reoccurs in more recent editions of this text.

References

Abernathy, W. and Utterback, J. (1978), “Patterns of industrial innovation”, Technology Review,Vol. 50, pp. 41-7.

Abrahamson, E. and Rosenkopf, L. (1997), “Social network effects on the extent of innovationdiffusion: a computer simulation”, Organization Science, Vol. 8, pp. 229-30.

Aggarwal, P., Cha, T. and Wilemon, D. (1998), “Barriers to the adoption of really-new productsand the role of surrogate buyers”, Journal of ConsumerMarketing, Vol. 15 No. 4, pp. 358-71.

Alba, J. and Hutchinson, J. (1987), “Dimensions of consumer expertise”, Journal of ConsumerResearch, Vol. 13, March, pp. 411-54.

Bauer, M. (1995), Resistance to New Technology, Cambridge University Press, Cambridge.

Bower, J. and Cristensen, C. (1995), “Disruptive technologies: catching the wave”, HarvardBusiness Review, January-February, pp. 43-53.

Brandenburger, A. and Nalebuff, B. (1995), “The right game. Use game theory to shape strategy”,Harvard Business Review, July-August, pp. 57-71.

Brandyberry, A. (2003), “Determinants of adoption for organisational innovations approachingsaturation”, European Journal of Innovation Management, Vol. 6 No. 3, pp. 150-8.

Brown, J. and Duguid, P. (1991), “Organizational learning and communities-of-practice: toward aunified view of working, learning, and innovation”, Organization Science, Vol. 2 No. 1,pp. 40-57.

Brown, S.L. and Eisenhardt, K.M. (1995), “Product development: past research, present findings,and future directions”, The Academy of Management Review, Vol. 20 No. 2, pp. 343-79.

Brucks, M. (1985), “The effects of product class knowledge on information search behavior”,Journal of Consumer Research, Vol. 12, pp. 1-16.

Bruland, K. (1995), “Patterns of resistance to new technologies in Scandinavia: an historicalperspective”, in Bauer, M. (Ed.), Resistance to New Technology, Cambridge UniversityPress, Cambridge.

Burt, R. (1987), “Social contagion and innovation: cohesion versus structural equivalence”,American Journal of Sociology, Vol. 92 No. 6, pp. 1287-335.

Burt, R. and Janicik, A. (1996), “Social contagion and social structure”, in Iacobucci, D. (Ed.),Networks in Marketing, Sage Publications, Thousand Oaks, CA, pp. 32-49.

Bussey, W. and Haigler, R. (1981), “Tubes versus transistors in electric guitar amplifiers”, IEEEInternational Conference on Acoustics, Speech, and Signal Processing, Vol. 6, pp. 800-3.

Calantone, R. and Montoya-Weiss, M. (1993), “Product launch and follow on”, in Souder, W.E.and Sherman, J.D. (Eds), Managing New Technology Development, McGraw-Hill, NewYork, NY, pp. 217-48.

Chatman, E. (1996), “The impoverished life-world of outsiders”, Journal of the American Societyfor Information Science, Vol. 47 No. 3, pp. 193-206.

Cohen, W. and Levinthal, D. (1990), “Absorptive capacity: a new perspective on learning andinnovation”, Administrative Science Quarterly, Vol. 35, pp. 128-52.

Coleman, J., Katz, E. and Menzel, H. (1966), Medical Innovation: A Diffusion Study, TheBobbs-Merrill Company, Indianapolis, IN.

EJIM13,2

218

Davis, F. (1989), “Perceived usefulness, perceived ease of use, and user acceptance of informationtechnology”, MIS Quarterly, September, pp. 319-40.

Dopfer, K., Foster, J. and Potts, J. (2004), “Micro-meso-macro”, Journal of Evolutionary Economics,Vol. 14, pp. 263-79.

Easingwood, C. and Koustelos, A. (2000), “Marketing high technology: preparation, targeting,positioning, execution”, Business Horizons, May-June, pp. 27-34.

Ellen, P., Bearden, W. and Subhash, S. (1991), “Resistance to technological innovations: anexamination of the role of self-efficacy and performance satisfaction”, Journal of theAcademy of Marketing Science, Vol. 19 No. 4, pp. 297-308.

Fornell, C. (1992), “A national customer satisfaction barometer: the Swedish experience”, Journalof Marketing, Vol. 56, January, pp. 6-21.

Galende, J. (2006), “Analysis of technological innovation from business economics andmanagement”, Technovation, Vol. 26, pp. 300-11.

Gandal, N. (2002), “Compatibility, standardization, and network effects: some policy.implications”, Oxford Review of Economic Policy, Vol. 18, pp. 80-91.

Gandal, N., Michael, K. and Rafael, R. (2000), “The dynamics of technological adoption inhardware/software systems: the case of compact disc players”, Rand Journal of Economics,Vol. 31, pp. 43-61.

Gilfillan, S.C. (1935), Inventing the Ship, Follett Publishing Co, Chicago, IL.

Haggman, S. (2009), “Functional actors and perceptions of innovation attributes: in?uence oninnovation adoption”, European Journal of Innovation Management, Vol. 12 No. 3,pp. 386-407.

Hall, B.H. and Khan, B. (2003), Adoption of New Technology, NBER Working Paper, N. 9730,NBER, Cambridge, MA.

Hanninen, S. and Sandberg, B. (2006), “Consumer learning roadmap: a necessary tool for newproducts”, International Journal of Knowledge and Learning, Vol. 2 Nos 3/4, pp. 298-307.

Henard, D. and Szymanski, D. (2001), “Why some new products are more successful than others”,Journal of Marketing Research, Vol. 38 No. 3, pp. 362-75.

Janis, I. (1972), Victims of Groupthink, Houghton Mifflin Company, Boston, MA.

Kim, S. (2009), “The integrative framework of technology use: an extension and test”, MISQuarterly, Vol. 33 No. 3, pp. 513-37.

Kingsley, P. and Anderson, T. (1998), “Facing life without the internet”, Internet Research:Electronic Networking Applications and Policy, Vol. 8 No. 4, pp. 303-12.

Kling, R. (1999), “Can the ‘next-generation internet’ effectively support “ordinary citizens”?”,The Information Society, Vol. 15, pp. 57-63.

Krieg, R. (1995), “Information technology and low-income, inner-city communities”, Journal ofUrban Technology, Vol. 3 No. 1, pp. 1-17.

Laukkanen, T., Sinkkonen, S., Kivijarvi, M. and Laukkanen, P. (2007), “Innovation resistanceamong mature consumers”, Journal of Consumer Marketing, Vol. 24 No. 7, pp. 419-27.

Lee, E., Kwon, K. and Schumann, D. (2005), “Segmenting the non-adopter category in thediffusion of internet banking”, International Journal of Bank Marketing, Vol. 23 No. 5,pp. 414-37.

Maidique, M. and Zirger, B. (1984), “A study of success and failure in product innovation: thecase of the US electronic industry”, Harvard Business Review, Vol. 61, May/June,pp. 192-203.


219

March, J. (1991), “Exploration and exploitation in organizational learning”, Organization Science,Vol. 2 No. 1, pp. 71-87.

Maryse, B. and Eelko, H. (2008), “Into the drivers of innovation adoption”, European Journal ofInnovation Management, Vol. 11 No. 1, pp. 5-24.

Miller, P. (1994), “Gender discrimination in training: an Australian perspective”, British Journalof Industrial Relations, Vol. 32 No. 4, pp. 539-63.

Montoya-Weiss, M. and Calantone, R. (1994), “Determinants of new product performance: areview and meta-analysis”, Journal of Product Innovation Management, Vol. 11 No. 5,pp. 397-417.

Moore, G. and Benbasat, I. (1991), “Development of an instrument to measure the perceptions ofadopting an information technology innovation”, Information Systems Research, Vol. 2No. 3, pp. 192-222.

Moreau, C., Lehmann, D. and Markman, A. (2001), “Entrenched knowledge structures andconsumer response to new products”, Journal of Marketing Research, Vol. 8, February,pp. 14-29.

Morris, M. and Venkatesh, V. (2000), “Age differences in technology adoption decisions:implications for a changing work force”, Personnel Psychology, Vol. 53, pp. 375-403.

Murmann, J.P. and Frenken, K. (2006), “Towards a systematic framework for research ondominant designs, technological innovations, and industrial change”, Research Policy,Vol. 35, pp. 925-52.

Nonaka, I. (1994), “A dynamic theory of organizational knowledge creation”, OrganizationScience, Vol. 5 No. 1, pp. 14-37.

Richins, M. and Bloch, P. (1986), “After the new wears off: the temporal context of productinvolvement”, Journal of Consumer Research, Vol. 13, September, pp. 280-5.

Rogers, E. (1962), Diffusion of Innovations, The Free Press, New York, NY.

Schiavone, F. and MacVaugh, J. (2009), “A user-based perspective on limits to the adoption ofnew technology”, International Journal of Technoentrepreneurship, Vol. 2 No. 2, pp. 99-114.

Seligman, L. (2006), “Sensemaking throughout adoption and the innovation-decision process”,European Journal of Innovation Management, Vol. 9 No. 1, pp. 108-20.

Selwyn, N. (2003), “Apart from technology: understanding people’s non-use of information andcommunication technologies in everyday life”, Technology in Society, Vol. 25, pp. 99-116.

Shapiro, C. and Varian, H.R. (1999), Information Rules: A Strategic Guide to the NetworkEconomy, Harvard Business School Press, Boston, MA.

Shy, O. (2001), The Economics of Network Industries, Cambridge University Press, Cambridge.

Slowlkowski, S. and Jarratt, D. (2007), “The impact of culture on the adoption of high technologyproducts”, Marketing Intelligence & Planning, Vol. 15 No. 2, pp. 97-105.

Smith, T. (2007), “Ten reasons why you should buy a mac”, available at: www.reghardware.co.uk/2007/03/21/ten_reasons_to_buy_a_mac/ (accessed 5 November 2009).

Song, M. and Montoya-Weiss, M. (1998), “Critical development activities for really new versusincremental products”, Journal of Product Innovation Management, Vol. 15, pp. 124-35.

Souder, W. and Sherman, J. (1993), Managing New Technology Development, McGraw-Hill, NewYork, NY.

Sterman, J. (2001), “System dynamics modeling: tools for learning in a complex world”, CaliforniaManagement Review, Vol. 43 No. 4, pp. 8-25.

Stuart, R. and Abetti, P. (1987), “Start-up ventures: towards the prediction of initial success”,Journal of Business Venturing, Vol. 2 No. 3, pp. 215-30.

EJIM13,2

220

Taylor, M.R., Rubin, E.S. and Hounshell, D.A. (2003), “Effect of government actions ontechnological innovation for SO2 control”, Environmental Science Technology, Vol. 37No. 20, pp. 4527-34.

Teece, D. (1986), “Profiting from technological innovation: implications for integration,collaboration, licensing and public policy”, Research Policy, Vol. 15, pp. 285-305.

Venkatesh, V., Speier, C. and Morris, M. (2003), “User acceptance of information technology:toward a unified view”, MIS Quarterly, Vol. 27 No. 3, pp. 425-77.

Wenger, E. (1998), “Communities of practice: learning as a social system”, The Systems Thinker,Vol. 9 No. 5.

Zeithmal, V. (1988), “Consumer perceptions of price, quality, and value: a means-end model andsynthesis of evidence”, Journal of Marketing, Vol. 52, July, pp. 2-22.

Further reading

Nelson, R. and Winter, S. (1982), An Evolutionary Theory of Economic Change, HarvardUniversity Press, Cambridge, MA.

Rosenberg, N. (1969), “The direction of technological change: inducement mechanisms andfocusing devices”, Economic Development and Cultural Change, Vol. 18, pp. 1-24.

Schilling, M. (2006), Strategic Management of Technological Innovations, 2nd ed., McGraw-Hill,New York, NY.

Schmookler, J. (1962), “Economic sources of inventive activity”, Journal of Economic History,Vol. 22, pp. 1-20.

Schumpeter, J. (1934), The Theory of Economic Development, Harvard University Press,Cambridge, MA.

Shannon, C. (1948), “A mathematical theory of communication”, Bell System Technical Journal,Vol. 27, July and October, pp. 379-423.

Urban, G. and Von Hippel, E. (1988), “Lead user analyses for the development of new industrialproducts”, Management Science, Vol. 34 No. 5, pp. 569-82.

Van Gigch, J. (1978), Applied General Systems Theory, Harper & Row, New York, NY.

Von Hippel, E. (1988), The Sources of Innovation, Oxford University Press, New York, NY.

Webster, F. (1995), Theories of the Information Society, Routledge, London.

Wejnert, J. (2002), “Integrating models of diffusion of innovations: a conceptual framework”,Annual Review of Sociology, Vol. 28, pp. 297-326.

Corresponding authorJason MacVaugh can be contacted at: [email protected]


221

To purchase reprints of this article please e-mail: [email protected] visit our web site for further details: www.emeraldinsight.com/reprints

Documents

Limits to the diffusion of innovation