M01_Afuah - Innovation Management. strategies, implementation, and profits.pdf

c H T E R MODELS OF INNOVATION One question has interested management scholars for years: what type of finn is most likely to innovate? In this chapter we explore some of the models that have been advanced in an effort to be...

c H T E R MODELS OF INNOVATION One question has interested management scholars for years: what type of finn is most likely to innovate? In this chapter we explore some of the models that have been advanced in an effort to better answer this question. This exploration pro- vides some background definitions and concepts that are crucial to an understand- ing of the main theme of this book: recognizing the potential of an innovation, profiting from it, and protecting those profits. We first define innovation and then describe each modeJ, pointing out its shortcomings and contributions. From these models we synthesize a framework forexpioring who is likely to introduce and ex- ploit innovation. DEFINITION OF INNOVATION Innovation is the use of new knowledge to offer a new product or service that cus- tomers want. It is invention + commercialization. I It is, according to Porter, "a new way of doing things (tenned invention by some authors) that is commen:ialiloed. The process of innovation cannot be separated from a finn's strategic and compet- itive context.''l The new knowledge) can be technological or market related. Tech- nological knowledge is knowledge of components, linkages between components, methods, processes, and techniques that go into a product or service. Market knowledge is knowledge of distribution channels, product applications, and cus- tomer expectations, preferences, needs, and wants. The product or service is new in that its cost is lower, its attributes are improved, it now has new attributes it never had before, or it never existed in that market before. Often the new product or ser- vice itself is called an innovation, reflecting the fact that it is the creation of new technological or market knowledge. Thus the discovery and development of Merck's cholesterol-bursting drug Mevacor was an innovation, as were Wal-Mart's location of discount retail stores in small towns in the southwestern United States and Federal Express's one-day delivery service. Innovation has also been defined as "the adoption of ideas that are new to the adopting organization."· 11lis book is about profiting from innovation. Generating good ideas or adopting a new one, in and of itself, is only a start. To be an innova- tion, an idea must be converted into a product or service that customers want Com-  14 F~ntals Models of inmJ\'(uiort IS ing up with the idea or prototypc---invention- is one thing. Championing it. shep- change in the firm's ability to offer a new product (Figure 2. 1). Thus innovation can herding it, and nurturing it into a product or service that customers want is another. be defined in terms of the extent to which it impacts a finn' s capabilities. This is Innovation entails both invention and commercialization. what is usually referred to as the organizational view (of classifying innovations). A distinction has also been made between technical and administrative ione- In this view, an innovation is said to be radical if the technological knowledge re- vation.j Technical innovation is about improved products, services, or processes quired to exploit it is very different from existing knowledge, rendering existing or completely new ones. This contrasts with administrative innovation. which knowledge obsolete. Such innovations are said to be competence destroying. II Re- pertains to organizational structure and administrative processes and mayor may frigerators were a radical innovation because making them required firms to inte- nol affecllechnical innovation. Technical innovation mayor may nOI requi re ad- grate knowledge of thermodynamics, coolants, and electric motors, which was ministrative innovation. A technical innovation can be a product or a process. Ac- very different from knowledge of harvesting and hauling ice. At the other end of cordi ng to Damanpour, product innovations "are new products or services intro- the dichotomy is incremental innovation. In it, the knowledge required to offer a duced to meet an external and market need," whereas process innovations are product builds on existing knowledge. II is, according to Thshman and Anderson, "new elements introduced into an organization's production or service opera- competence enhancing. For example, a "shrink" of Intel's Pentium chip to make it tions-input materials, task specifications, work and information flow mecha- run at 200 MHz is an incremental innovation in the organizational sense since the nisms, and equipment used to produce a product or render a service.'" Thus, 10 knowledge required to do so builds on the finn 's knowledge in microprocessor de- the extcnt that technical innovations use new technological or market knowledge velopment. Most innovations are incremental. to offer new products or services to customers, they fit the definition of innova- In the second place, since innovation results in superior products (lower cost, tion in this book. better or new features), it can also be classified as a function of the extent to which it renders old products noncompetitive. This is the so-called economic (competi- tiveness) view. In this view, an innovation is said to be radical (drastic) if it results in a product that is so superior (lower cost. better attributes, or new attributes) that WHO INNOVATES? existing products are rendered noncompetitive, For example, the mechanical cash The debate over who is most likely to innovate dates back to, at least, Schumpeter, register could not compete with electronic point-Qf-sale systems (EPOSs), Thus who first suggested that small entrepreneuria1 firm s were the sources of most innl)- EPOSs were a radical innovation in the economic sense, Very often, however, the vations, 1 Later he changed his view and suggested that, for several reaso ns. large innovation still allows existing products to stay competitive. In that case it is said to finns with some degree of monopoly power were more likely to be the sources of be incremental or nondrastic, Both diet and caffeine-free sodas are incremental in- technological innovation. He argued thai large firms have the production and other complementary assets that are necessary to commercialize an invention: have the size to exploit the economies of scale that are prevalent in R&D; are more diversi- fied and therefore 'more willing to take the kind of risk that is inherent in R&D proj- New product ] ects; have better access to capital than smaller finns; and, as monopolists, do not 'Lo.. 00It have competitors ready to imitate their innovations and therefore are more likely to tmprovecl Ittributu invest in them.' Empirical studies in search of support for either position have not Ne.. anrib\rla been able to establish a clear relationship between a finn's size and market power and its innovative activity.' By shifting the focus 10 the type of innovation, however, some research suggesl.S that whether incumbents or new entrants are able to intrl)- t duce and exploit innovation is a function of whether the innovation is incremental Competences or radical, thaI is, a function o f how new the new knowledge and the new product wd (in our definition of innovation) are,IO ~~" STATIC MODELS N~ t Now Incremental versus Radical Innovation technological m,rl", knowledge knowledge An innovation has two kinds of impact on a finn , In the first place, since knowledge underpins a finn 's ability to otTer products, a change in knowledge implies a FIGURE 2, I, fnnovation: Organi7.ational and economic im plications. 16 Modtls of inno't'tlti.on 17 novations in the economic sense because their introduction allowed the standard Innovation colas to remain competitive. Inc~ment.l Radical These organizational and competitiveness definitions of incremental and radi- GB_~\llinlhe cal innovation serve as the basis for two views of who is likely to innovate: the Mo.let and lIIkllog IhE:ir ~rromX"')'IIOCAT strattgic incentive (to invest) and the organizational capabilities views. lncllqoy market IeIdmhIp _»MRI Incumbent pn;6onI ill 1M ~Iion froaI 64K 10 '256K DRAM IBM auo::eufllllJ mD!he Strategic i ncentive (to [nvest) trl/U.llMlfts trom YlOIUm lubes 10 u.t1slOl'l1O IDfepucd cirruilJ In In the strategic incentive to invest view, the type of innovation-whether it is radi- malnframe. computen cal or incremental in the economic sense-determines what type of firm is likely 10 invest to be the first to innovate. Given that a radical innovation renders an incum- bent's existing products noncompetitive, an incumbent with market power may be New entrant reluctant to invest in the innovation for fear of cannibalizing its existing products if, in doing so, it advances the date of introduction of the innovation. II New entrants, on the other hand, have less to lose. If they do not invest in the innovation, they have no products to sell in that market. On the other hand, incumbents would be more w illing 10 invest in incremental innovations since such investments allow FIGURE 2.2. COWlterexamples to predictions by incremental and radical models. Ihei r ex isting products to stay competitive in the markel. ll One major shortcoming of this model is thai it assumes that firms have reeog· nired the potential of the innovation, and in the case of a radical innovation, that only the fear of cannibalization prevents them from exploiting it This is not always ter when the innovation is incremental. In many industries, however, incumbents the case, especially with radical innovations. In any case, the model may explain have been the first to introduce or exploit radical innovations and, in some cases, why some new entrants have been the fi rst to embrace radical innovation and in· have fai led to exploit incremental innovations. Figure 2.2 gives some examples. cumbents have been responsible for the bulk of incremental innovations. II does Intel and Mostek lost their leadership positions in DRAM memory chips during the not, however, explai n why incumbents who invest in radical innovation may still transition from the 64K to the 256K chip, an incremental innovation in the organi- fail. This is where the organizational capabilities view comes in. zational sense, despite being very strong incumbents and investing heavily in the 256K DRAM. On the other band, GE, an incumbent in the diagnostic medical Organizationa l Capabilities equipment industry, maintained its leadership position in the transition from X rays to CAT scans to MRI, relatively radical innovations. mM was able to maintain its If an innovation is radical in the organizational sense, incumbents have two prob· domi nant position in the computer industry during the radical transitions from vac- lems in exploiting it. In the first place, since the change is competence destroying, uum tubes to transistors to integrated circuits. '4 they do not have the capabilities to exploit it. Ln the second place, and perhaps The counterexamples illustrated in Figure 2.2 suggest that we need more than more important, the firm 's existing capabilities may not only be useless, they may the incremental-radical model to predict the outcome of technological change. We actual ly be a handicap to the introduction and development of the innovation." need more models that explain why some incumbents are the first 10 embrace or ex· Firms find it difficult to break their habits, the routines and procedures they had put ploit radical technological change and why they sometimes fail to exploit some in- in place to exploit the old technology. They must unlearn the old ways of doing cremental innovations. things. New entrants, on the other hand, do not have the burden of the old technol· ogy and can go on unencumbered to build capabilities for the innovation and ex- ploit it. If, on the other hand, the innovation is incremental, incumbents tend to domi- Abema thy-Clark Model nate since the required knowledge builds on what they ab'eady have, but new en- trants would have 10 build it from scratch. The Abemathy-Clark model offers one explanation why incumbents may oUlper· I' fonn new entrants in the face of some "radical" innovations. The model suggests that there are actually two kinds of knowledge that underpin an innovation: techno- Issues with the Incremental-Radical Dichotomy logical and market. Thus a finn 's technological capabilities could become obsolete One implication of these definitions is that new entrants are more likely to do well while its market capabilities remain intact. If such market capabilities are impor- in Ihe face of a radical innovation, whereas incumbents are more likely to fare bet- tant and diffi cult to acquire, an incumbent whose technological capabilities have 18 Fwtdam~nlals Mod~ls oj iMOvalioll.. Tedtnleal capabiUtiel Arthittctunl knowl«lge Preserved Dest royed Enhanced Destroyed ReguIu Revolutionary J Enh,",," Incremental Architectural ~ I ~--~----~ ~ Destroyed Niche Architectural Ii!.:s ""troy'" Modular Radical FIGURE 2,3, Abematby-Clan: (mninoiogy-roie of technological FIGURE 2.4. Architectural illll(Mllion. and market capabilities. Source: Reprinted from "ArchiteclUl'lIi innovation: The reconfigUl1ll- tion of existing product ~hnologies and the failure of established firms" by R. Henderson and K. B. Clark, published in been destroyed can use the market ones to its advantage over a new entrant. Focus- Administmti\le Science Quarterly 35( 1). March 1990, by ing on the perspective of the innovating finn. the model classifies innovations ac- pennission of Administrative Science Quaner]y. Copyright C 1990 by Cornell University. cording to their impact on the existing technological and market knowledge of the manufacturer (Figure 2.3). An innovation is regular if it conserves the manufac- turer's existing technological and market capabilities, niche if it conserves techno- logical capabilities but obsole.tes market capabilities, revolutionary if it obsoletes nent knowledge or architectural knowledge, or both, with different consequences technological capabilities but enhances market capabilities, and architectural if for the finn adopting it. They went on to define four kinds of innovations, as illus- both technological and market capabilities become obsolete. trated in Figure 2.4. If the innovation enhances both component and architectural While these different categories of innovation are illuminating, the point to knowledge, it is incremental; if it destroys both component and architectural note in this model is that market knowledge can be just as important as technolog- knowledge, it is radical. However, if only the architectural knowledge is destroyed ical knowledge. For example, GE's market capabilities were instrumental to its and the component knowledge enhanced, the innovation is architectural. The last perfonnance in the transition from one generation of technologically radical inno- case, where component knowledge is destroyed but architectural knowledge en- vation to another in the medical diagnostics equipment industry.!' It was not the hanced, is called modular innovation. first to introduce each new technology, but it successfull y transitioned from X rays With these definitions it became clear why finns had problems with what ap- to CAT scans to MRl, all technologically competence-destroying innovations. peared to be incremental innovation. They may have mistaken architectural inno- vation for incremental innovation. While the component knowledge required to ex- ploit the innovations had not changed (and therefore the semblance of incremental innovation), architectural knowledge had changed. Architectural knowledge is Henderson-Clark Mode l often tacit and embedded in the routines and procedures of an organization, making Henderson and Clark were puzzled why some incumbents have so much difficulty changes in it difficult to discern and respond to. in dealing with what appear to be "incremental" innovations-seemingly minute changes in existing technologies. Xerox stumbled for many years before finally de- veloping a good small plain-paper copier, despite being the pioneer of the core Disruptive Technological Change Model technology of xerograpby.1I RCA was never able to lead in the market for portable transistor radios despite its experience in the components (transistors, audio ampli- According to the disruptive technological change model, advanced by Professor fiers, and loudspeakers) that went into the portable radio. From their research, Hen- Clayton Christensen, incumbents fai l to exploit disruptive technologies not so much derson and Clark suggested that since products are nonnally made up of compo- because these finns do not "get It," as suggested by the architectural innovation nents connected together, building them must require two kinds of knowledge: model, or because the technologies arc competence destroying to them, as sug- knowledge of the components and knowledge of the linkages between them, which gested by the incremental-radical model. 19 Rather, incumbents fail because they they call architectural knowledge. An innovation, then, can impact either compo- spend too much time listening to and meeting the needs of their existin2 mainstream 20 F~fltals Modtls of I lIIIiwaliorl customers who, initially. have no use for products from the disruptive technology. meet the perfonnance requirements demanded by B. The finn that has produced " Disruptive technologies have the following four characteristics: A over the years and has been successful at doing so focuses its attention on sat- isfying the requirements of its key ex.isting customers and therefore does not pay I. They create new markets by intnXlucing a new kind of product or service. attention to developing the necessary capabilities. processes, and culture to build 2. The new product or service from the new technology costs less than exist- product C. which meets the performance attributes 0 that are needed by a differ- ing products or services from the old technology. ent market. Start-ups and other entrants produce C and keep improving its per- 3. Initially, the products perfonn worse than existing products when judged ronnance. In Year 5, say. C's performance improves to a point where it now by the performance metrics that mainstream existing customers value. meets the needs of the market with del;lland B-the finn 's mainstream customers. Eventually, however, the perfonnance catches up and addresses the needs Because C is lower in cost than A, the finn 's mainstream customers abandon it of mainstream customers. for the stan-ups who offer C. At this time. it is too late for the finn and other pro- 4. The technology should be difficull to protect using patents. ducers of A to quickly shed their processes, val ues, and culture that served them so well with the old technology (in developing A) to develop C and gain a prod- To understand the disruptive technological change model, consider a firm that uct advantage. Stan-ups and other entrants take the leadership position in pro- has been successful in exploiting an existing technology to offer products to its ducing C. mainstream customers. The firm' s capabilities-what it can or cannot do-are a functio n of its resources, processes, and values. Its resources are assets such as product designs, brands, relationships with suppliers, customers, distribution. Innovation Value-Added Chain people, plants and equi pment. technologies, and cash reserves. Its processes are the systems that the firm has put in place to transform resources into bener cus- The innovation value-added chain20 model can explain both why an incumbent can tomer value. These systems are designed to make task performance more effi - outperform new entrants at radical innovation and why it may also fail at incre- cient and are difficu lt to change, especially when they have been embedded into mental innovation. It differs from previous models in that while these other models organizational culture. Suppose A is one of the products that the finn develops focus on the impact of innovation on a firm 's capabilities and competitiveness, it using an existing technology. and thaI, in Year 2. more than meets the key perfor- focuses on what the innovation does to the competitiveness and capabilities of a mance attributes that the finn 's mainstream customers want (B) in the product firm 's suppliers. customers, and complementary innovators. 21 That is, previous (Figure 2.5). Also suppose that in Year 2, a new product C, which costs less than models addressed the question: what does the electric car do to Ford's technologi- A, is introduced. Initially, C's performance is inferior to that of A and does not cal and market capabilities? But innovation also has implicatio ns for suppliers, customers, and complementary innovators, implications that can have far-reaching effects on the manufacturer.21 Thus the electric car has implications no{ only for Ford's technical and market knowledge. but also for that of its suppliers (e.g., sup- pliers of electronic fue l inj ection systems), customers, and complementary innova- -- tors, such as gas station owners and oil companies. An innovation that is incremen- tal to a manufacturer can be radical to its customers and complementary innovators and incremental to its suppliers. For example, the DSK (Dvorak simplified key- board) keyboard arrangement that by some estimates performs 20--40 percent bet- Measure of key ter than the QWERTY arrangement, which most of today 's keyboards have, was perfonnance attribute -- competence enhancing to its innovator, Dvorak, and other typewriter manufactur- ers.u All they had to do was rearrange the position of the keys if they wanted to manufacture the DSK. But it was competence destroying to customers who had al- -- ready learned how to type with the QWERTY keyboard since to use the new key- board, they would have to relearn how to touch-type again. The various phascs of this innovation at the different stages of the innovation value-added chain are shown in Figure 2.6. 2 4 , Another example (also ilJustrated in Figure 2.6) is eray Computer's decision in 1988 to develop and market a supercomputer that would use gallium arsenide (GaAs)2' c;:hips- a technology that yields very fast chips and consumes very little power, but was still relatively unproven then-instead of the proven silicon chip AGUItE 2.5. Disruptive technological change. technology in which its suppliers had built their competences. While the super-  Mod~/s of i1lllOI'(lfion 2J F~nJa/s " c.g > 0 c ""'" El«1ric I;C H B. H - - -........ -... "-'.- /........ ~ / Hu-- -........... ==-.5 '0 / 0.- / 0-- Il. c.., ,- , R~lnr_ -.., -- (: c.._ incm!ll'RW Ii """""....,.... =-- / \ """-'./ /.1 1 ! a !> J !! c.._./ ,-"""'"" -/./ :i Value-added chain {! 8.! I u_... _. IL -....... -- / ~- ~ Ii ~ FIGURE 1.6. Innovation value-added chain. An innovation has implications not only -=~ for the firm but also for ilS suppliers. customers. and complementary innovators. c....._. / / '- computer design was competence enhancing to Cray, its decision 10 use gallium ar- 1(-....... - "'"-'........ - - senide was competence destn;lying to its lJ'aditional silicon chip supplier base..... --;,.... ,- i! "'-'................ V In each of these examples, the innovation may have a different impact al each of / -- -....... ~ the stages of the innovation value-added chai n, suggesting that an innovation that is incremental to the manufacturer may not be to supplicn , customers, or comple- U - RanfOlCed o.~_ mentary innovators. Thus incumbents for whom an innovation is competence de- c.._ stroying may still do well if the innovation is competence enhancing to their value chain, and relations with the chain are important and difficult to establish. The im- ""'................. / ~n plications are thaI a firm' s success in exploiting an innovation may depend as much M.~I"(_ 0.."""""" on whatlhe innovation does to Ihe capabilities pf Ihe finn as 01\ what it does to Ihe 0......... capabilities of its innovation value-added chain of suppliers, customers, and com- FIGURE 2.7. Imp&l.:t oftn innovation on an innovation value-added chain. plementary innovators (Figure 2.7). SoUI'CtI: RqJrinltld ITom Rutlal'Cll Policy 2