Explicating dynamic capabilities: the nature and ...pages.stern.nyu.edu/~eyoon/seminar/CrossDisciplinary/...Explicating Dynamic Capabilities: Nature and Microfoundations must the innovating

Strategic Management JournalStrat. Mgmt. J. (2007)

Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/smj.640

Received 16 February 2004; Final revision received 20 June 2007

EXPLICATING DYNAMIC CAPABILITIES: THE NATUREAND MICROFOUNDATIONS OF (SUSTAINABLE)ENTERPRISE PERFORMANCE

DAVID J. TEECE*Institute of Management, Innovation and Organization, Haas School of Business,University of California, Berkeley, California, U.S.A.

This paper draws on the social and behavioral sciences in an endeavor to specify the natureand microfoundations of the capabilities necessary to sustain superior enterprise performancein an open economy with rapid innovation and globally dispersed sources of invention, innova-tion, and manufacturing capability. Dynamic capabilities enable business enterprises to create,deploy, and protect the intangible assets that support superior long- run business performance.The microfoundations of dynamic capabilities—the distinct skills, processes, procedures, orga-nizational structures, decision rules, and disciplines—which undergird enterprise-level sensing,seizing, and reconfiguring capacities are difficult to develop and deploy. Enterprises with strongdynamic capabilities are intensely entrepreneurial. They not only adapt to business ecosystems,but also shape them through innovation and through collaboration with other enterprises, enti-ties, and institutions. The framework advanced can help scholars understand the foundations oflong-run enterprise success while helping managers delineate relevant strategic considerationsand the priorities they must adopt to enhance enterprise performance and escape the zero profittendency associated with operating in markets open to global competition. Copyright 2007John Wiley & Sons, Ltd.

INTRODUCTION

Recent scholarship stresses that business enter-prises consist of portfolios of idiosyncratic anddifficult-to-trade assets and competencies (’re-sources’).1 Within this framework, competitiveadvantage can flow at a point in time from theownership of scarce but relevant and difficult-to-imitate assets, especially know-how. However, in

Keywords: cospecialization; intangible assets; innovation;business ecosystems; entrepreneurship; managerial capi-talism; global competitiveness*Correspondence to: David J. Teece, F402 Haas School ofBusiness #1930, University of California, Berkeley, California94720-1930, U.S.A. E-mail: [email protected] The reference here is to the resource-based theory of theenterprise advanced by Rumelt (1984), Wernerfelt (1984), Amitand Schoemaker (1993), and others. Some of my earlier work(Teece, 1980, 1982) was also in this vein.

fast-moving business environments open to globalcompetition, and characterized by dispersion in thegeographical and organizational sources of inno-vation and manufacturing, sustainable advantagerequires more than the ownership of difficult-to-replicate (knowledge) assets. It also requiresunique and difficult-to-replicate dynamic capabili-ties. These capabilities can be harnessed to con-tinuously create, extend, upgrade, protect, andkeep relevant the enterprise’s unique asset base.For analytical purposes, dynamic capabilities canbe disaggregated into the capacity (1) to senseand shape opportunities and threats, (2) to seizeopportunities, and (3) to maintain competitivenessthrough enhancing, combining, protecting, and,when necessary, reconfiguring the business enter-prise’s intangible and tangible assets. Dynamiccapabilities include difficult-to-replicate enterprise

Copyright 2007 John Wiley & Sons, Ltd.

D. J. Teece

capabilities required to adapt to changing cus-tomer and technological opportunities. They alsoembrace the enterprise’s capacity to shape theecosystem it occupies, develop new products andprocesses, and design and implement viable busi-ness models. It is hypothesized that excellencein these ‘orchestration’2 capacities undergirds anenterprise’s capacity to successfully innovate andcapture sufficient value to deliver superior long-term financial performance. The thesis advanced isthat while the long-run performance of the enter-prise is determined in some measure by how the(external) business environment rewards its her-itage, the development and exercise of (internal)dynamic capabilities lies at the core of enterprisesuccess (and failure). This paper first describes thenature of dynamic capabilities, and then explicatestheir microfoundations.

The ambition of the dynamic capabilities frame-work is nothing less than to explain the sources ofenterprise-level competitive advantage over time,and provide guidance to managers for avoiding thezero profit condition that results when homoge-neous firms compete in perfectly competitive mar-kets. A framework, like a model, abstracts fromreality. It endeavors to identify classes of relevantvariables and their interrelationships. A frameworkis less rigorous than a model as it is sometimesagnostic about the particular form of the theoreti-cal relationships that may exist. Early statements ofthe dynamic capabilities framework can be foundin Teece, Pisano, and Shuen (1990a, 1990b, 1997)and Teece and Pisano (1994). An extensive lit-erature on dynamic capabilities now exists (e.g.,Helfat et al., 2007) that can be organized and inte-grated into the general framework offered here.

As indicated, the possession of dynamic capabil-ities is especially relevant to multinational enter-prise performance in business environments thatdisplay certain characteristics. The first is that theenvironment is open to international commerce andfully exposed to the opportunities and threats asso-ciated with rapid technological change. The sec-ond is that technical change itself is systemic in

2 The management functions identified are analogous to that ofan orchestra conductor, although in the business context the‘instruments’ (assets) are themselves constantly being created,renovated, and/or replaced. Moreover, completely new instru-ments appear with some frequency, and old ones need to beabandoned. While flexibility is certainly an element of orches-tration, the latter concept implies much more.

that multiple inventions must be combined to cre-ate products and/or services that address customerneeds. The third is that there are well-developedglobal markets for the exchange of (component)goods and services; and the fourth is that the busi-ness environment is characterized by poorly devel-oped markets in which to exchange technologicaland managerial know-how. These characteristicscan be found in large sectors of the global econ-omy and especially in high-technology sectors. Insuch sectors, the foundations of enterprise successtoday depend very little on the enterprise’s abil-ity to engage in (textbook) optimization againstknown constraints, or capturing scale economiesin production. Rather, enterprise success dependsupon the discovery and development of opportuni-ties; the effective combination of internally gener-ated and externally generated inventions; efficientand effective technology transfer inside the enter-prise and between and amongst enterprises; theprotection of intellectual property; the upgradingof ‘best practice’ business processes; the inven-tion of new business models; making unbiaseddecisions; and achieving protection against imita-tion and other forms of replication by rivals. Italso involves shaping new ‘rules of the game’ inthe global marketplace. The traditional elementsof business success—maintaining incentive align-ment, owning tangible assets, controlling costs,maintaining quality, ‘optimizing’ inventories—arenecessary but they are unlikely to be sufficient forsustained superior enterprise performance.

Executives seem to recognize new challengesin today’s globally competitive environments andunderstand how technological innovation is nec-essary but not sufficient for success. A. J. Lafley,CEO of Proctor & Gamble, notes that ‘the nameof the game is innovation. We work really hard totry to turn innovation into a strategy and a process. . . ‘.3 Sam Pamisano, CEO of IBM, remarks that‘innovation is about much more than new prod-ucts. It is about reinventing business processes andbuilding entirely new markets that meet untappedcustomer demand.’4 Put differently, there is anemerging recognition by managers themselves thatthe foundations of enterprise success transcendsimply being productive at R&D, achieving newproduct introductions, adopting best practice, anddelivering quality products and services. Not only

3 Fortune, December 11, 2006: 4.4 Business Week, April 24, 2004: 64.

Copyright 2007 John Wiley & Sons, Ltd. Strat. Mgmt. J., (2007)DOI: 10.1002/smj

Explicating Dynamic Capabilities: Nature and Microfoundations

must the innovating enterprise spend heavily onR&D and assiduously develop and protect its intel-lectual property; it must also generate and imple-ment the complementary organizational and man-agerial innovations needed to achieve and sustaincompetitiveness.

As indicated, not all enterprise-level responsesto opportunities and threats are manifestations ofdynamic capabilities. As Sidney Winter (2003:991) notes, ‘ad hoc problem solving’ isn’t neces-sarily a capability. Nor is the adoption of a well-understood and replicable ‘best’ practice likely toconstitute a dynamic capability. Implementing bestpractice may help an enterprise become or remainviable, but best practices that are already widelyadopted cannot by themselves in a competitivemarket situation enable an enterprise to earn morethan its cost of capital, or outperform its competi-tors. Likewise, invention and innovation by them-selves are insufficient to generate success (Teece,1986).

Two yardsticks can be proposed for calibratingcapabilities: ‘technical’ fitness and ‘evolutionary’fitness (Helfat et al., 2007). Technical fitness isdefined by how effectively a capability performsits function, regardless of how well the capabilityenables a firm to make a living. Evolutionary orexternal fitness refers to how well the capabilityenables a firm to make a living. Evolutionary fit-ness references the selection environment. Helfatet al. (2007) further note that both technical andevolutionary fitness range from zero to some pos-itive value. These yardsticks are consistent withthe discussion here. Dynamic capabilities assist inachieving evolutionary fitness, in part by helpingto shape the environment. The element of dynamiccapabilities that involves shaping (and not justadapting to) the environment is entrepreneurial innature. Arguably, entrepreneurial fitness ought tohave equal standing with evolutionary fitness.

Dynamic capabilities have no doubt been rele-vant to achieving competitive advantage for sometime. However, their importance is now ampli-fied because the global economy has become moreopen and the sources of invention, innovation, andmanufacturing are more diverse geographicallyand organizationally (Teece, 2000), and multipleinventions must be combined to achieve market-place success (Somaya and Teece, 2007). Achiev-ing evolutionary fitness is harder today than it wasbefore the millennium. Moreover, regulatory andinstitutional structures must often be reshaped for

new markets to emerge; and as discussed later, theubiquity of ‘platforms’ must now be recognized(Evans, Hagiu, and Schmalensee, 2006).

While the development and astute managementof intangible assets/intellectual capital is increas-ingly recognized as central to sustained enter-prise competitiveness, the understanding of whyand how intangibles are now so critical stillremains opaque and is not addressed by orthodoxframeworks. What is needed is a new frameworkfor business and economic analysis. As formerU.S. Federal Reserve Chairman Alan Greenspanremarked, ‘we must begin the important work ofdeveloping a framework capable of analyzing thegrowth of an economy increasingly dominated byconceptual products.’5 The dynamic capabilitiesapproach developed here endeavors to be respon-sive to this challenge at the enterprise level.

In an earlier treatment (Teece et al., 1997: 530)it was noted that ‘we have merely sketched an out-line for a dynamic capabilities approach.’ In whatfollows, the nature of various classes of dynamiccapabilities is identified, and an effort is made toseparate the microfoundations of dynamic capa-bilities from the capability itself. Put differently,important distinctions are made between the orga-nizational and managerial processes, procedures,systems, and structures that undergird each classof capability, and the capability itself. One shouldnote that the identification of the microfounda-tions of dynamic capabilities must be necessarilyincomplete, inchoate, and somewhat opaque and/ortheir implementation must be rather difficult. Oth-erwise sustainable competitive advantage woulderode with the effective communication and appli-cation of dynamic capability concepts.

Of course, the existence of processes, proce-dures, systems, and structures already ubiquitouslyadopted by competitors does not imply that thesehave not in the past been the source of competitiveadvantage, or might not still be a source of compet-itive advantage in certain contexts. For example,studies of the diffusion of organizational innova-tions (e.g., Armour and Teece, 1978; Teece, 1980)

5 Chairman Alan Greenspan also noted recently, ‘over the pasthalf century, the increase in the value of raw materials hasaccounted for only a fraction of the overall growth of U.S. grossdomestic product (GDP). The rest of that growth reflects theembodiment of ideas in products and services that consumersvalue. This shift of emphasis from physical materials to ideas asthe core of value creation appears to have accelerated in recentdecades.’ (Remarks of Alan Greenspan, Stanford Institute forEconomic Policy Research, 2004.)


D. J. Teece

indicate that diffusion is by no means instanta-neous, and that profits can persist for many yearsbefore being competed away. Decade-long adop-tion cycles for new business structures and pro-cedures (e.g., performance measurement systems)are not uncommon. Uncertain imitability (Lippmanand Rumelt, 1982) may also serve to slow the dif-fusion process and support persistent differentialperformance.

Fortunately, the existing literature on strategy,innovation, and organization and the new literatureon dynamic capabilities have identified a panoplyof processes and routines that can be recognizedas providing certain microfoundations for dynamiccapabilities. For instance, Eisenhardt and Martin(2000) identify cross-functional R&D teams, newproduct development routines, quality control rou-tines, and technology transfer and/or knowledgetransfer routines, and certain performance mea-surement systems as important elements (micro-foundations) of dynamic capabilities. The efforthere is not designed to be comprehensive, but tointegrate the strategy and innovation literature andprovide an umbrella framework that highlights themost critical capabilities management needs to sus-tain the evolutionary and entrepreneurial fitness ofthe business enterprise.

SENSING (AND SHAPING)OPPORTUNITIES AND THREATS

Nature of the capability

In fast-paced, globally competitive environments,consumer needs, technological opportunities, andcompetitor activity are constantly in a state of flux.Opportunities open up for both newcomers andincumbents, putting the profit streams of incum-bent enterprises at risk. As discussed in Teeceet al. (1997), some emerging marketplace trajecto-ries are easily recognized. In microelectronics thismight include miniaturization, greater chip density,and compression and digitization in informationand communication technology. However, mostemerging trajectories are hard to discern. Sensing(and shaping) new opportunities is very much ascanning, creation, learning, and interpretive activ-ity. Investment in research and related activities isusually a necessary complement to this activity.

Opportunities get detected by the enterprisebecause of two classes of factors. First, as stressed

by Kirzner (1973), entrepreneurs can have differ-ential access to existing information. Second, newinformation and new knowledge (exogenous orendogenous) can create opportunities, as empha-sized by Schumpeter (1934). Kirzner stressed howthe entrepreneurial function recognizes any dise-quilibrium and takes advantage of it. The Kirzner-ian view is that entrepreneurship is the mechanismby which the economy moves back toward equi-librium. Schumpeter, on the other hand, stressedupsetting the equilibrium. As Baumol (2006: 4)notes, ‘the job of Schumpeter’s entrepreneur isto destroy all equilibria, while Kirzner’s worksto restore them. This is the mechanism under-lying continuous industrial evolution and revolu-tion.’ Equilibrium is rarely if ever achieved (Shane,2003). Both forces are relevant in today’s econ-omy.

To identify and shape opportunities, enterprisesmust constantly scan, search, and explore acrosstechnologies and markets, both ‘local’ and ‘dis-tant’ (March and Simon, 1958; Nelson and Winter,1982). This activity not only involves investmentin research activity and the probing and reprob-ing of customer needs and technological possibili-ties; it also involves understanding latent demand,the structural evolution of industries and mar-kets, and likely supplier and competitor responses.To the extent that business enterprises can openup technological opportunities (through engagingin R&D and through tapping into the researchoutput of others) while simultaneously learningabout customer needs, they have a broad menuof commercialization opportunities. Overcoming anarrow search horizon is extremely difficult andcostly for management teams tied to establishedproblem-solving competences. Henderson (1994)notes that General Motors (GM), IBM, and Dig-ital Equipment Corporation (DEC) encountereddifficulties because they became prisoners of thedeeply ingrained assumptions, information filters,and problem-solving strategies that made up theirworld views, turning the solutions that once madethem great into strategic straitjackets.

When opportunities are first glimpsed, entrepre-neurs and managers must figure out how to inter-pret new events and developments, which tech-nologies to pursue, and which market segmentsto target. They must assess how technologies willevolve and how and when competitors, suppli-ers, and customers will respond. Competitors mayor may not see the opportunity, and even if they



do they may calibrate it differently. Their actions,along with those of customers, suppliers, standard-setting bodies, and governments, can also changethe nature of the opportunity and the manner inwhich competition will unfold.

There are also constraints on the rules by whichcompetitive forces will play out. These constraintsare imposed by regulators, standard-setting bod-ies, laws, social mores, and business ethics. Theshape of the ‘rules of the game’ is thus theresult of co-evolution and complex interactionbetween what might be thought of as (business)ecosystem participants. Because of uncertainty,entrepreneurs/managers must make informed con-jectures about the path ahead. These conjecturesbecome working hypotheses that can be updated asevidence emerges. Once a new evolutionary pathbecomes apparent, quick action is needed.

Microfoundations

The literature on entrepreneurship emphasizes thatopportunity discovery and creation can originatefrom the cognitive and creative (’right brain’)capacities of individual(s). However, discovery canalso be grounded in organizational processes, suchas research and development activity. The abilityto create and/or sense opportunities is clearly notuniformly distributed amongst individuals or enter-prises. Opportunity creation and/or discovery byindividuals require both access to information andthe ability to recognize, sense, and shape devel-opments. The ability to recognize opportunitiesdepends in part on the individual’s capabilities andextant knowledge (or the knowledge and learningcapacities of the organization to which the indi-vidual belongs) particularly about user needs inrelationship to existing as well as novel solutions.This requires specific knowledge, creative activity,and the ability to understand user/customer deci-sion making, and practical wisdom (Nonaka andToyama, 2007). It involves interpreting availableinformation in whatever form it appears—a chart,a picture, a conversation at a trade show, news ofscientific and technological breakthroughs, or theangst expressed by a frustrated customer. One mustaccumulate and then filter information from profes-sional and social contacts to create a conjecture ora hypothesis about the likely evolution of technolo-gies, customer needs, and marketplace responses.This task involves scanning and monitoring inter-nal and external technological developments and

assessing customer needs, expressed and latent.It involves learning, interpretation, and creativeactivity.

While certain individuals in the enterprise mayhave the necessary cognitive and creative skills,the more desirable approach is to embed scan-ning, interpretative, and creative processes insidethe enterprise itself. The enterprise will be vulner-able if the sensing, creative, and learning functionsare left to the cognitive traits of a few individuals.6

Organizational processes can be put in place insidethe enterprise to garner new technical information,tap developments in exogenous science, monitorcustomer needs and competitor activity, and shapenew products and processes opportunities. Infor-mation must be filtered, and must flow to thosecapable of making sense of it. Internal argumentand discussion about changing market and tech-nological reality can be both inductive and deduc-tive. Hypothesis development, hypothesis ‘testing,’and synthesis about the meaning of informationobtained via search are critical functions, and mustbe performed by the top management team. Therigorous assembly of data, facts, and anecdotes canhelp test beliefs. Once a synthesis of the evidenceis achieved, recurrent synthesis and updating canbe embedded in business processes designed bymiddle management and/or the planning unit inthe business organization (Casson, 1997). If enter-prises fail to engage in such activities, they won’tbe able to assess market and technological devel-opments and spot opportunities. As a consequence,they will likely miss opportunities visible to others.

As noted in Teece et al. (1997), more decen-tralized organizations with greater local autonomyare less likely to be blindsided by market andtechnological developments. Because of the prob-lem of information decay as information movesup (and down) a hierarchy, businesses must devisemechanisms and procedures to keep managementinformed. Bill Hewlett and David Packard devel-oped ‘management by walking about’ (Packard,1995) as a mechanism to prevent top managementat Hewlett-Packard from becoming isolated from

6 In a limited sense, that is about decision making under uncer-tainty. As Knight observes, with uncertainty there is ‘a necessityto act upon opinion rather than knowledge’ (Knight, 1921: 268).The problem is not just about knowledge asymmetries and incen-tive problems as Alchian and Demsetz (1972) seem to suggest.Rather, it involves filtering and interpreting information aboutevolving technologies and marketplaces.


D. J. Teece

what was going on at lower levels in the enter-prise, and outside the enterprise as well. In otherorganizations (e.g., professional services) the man-agement ranks can be filled by leading profession-als who remain involved with professional work.This protects them from the hazards of managerialisolation.

The search activities that are relevant to ‘sens-ing’ include information about what’s going on inthe business ecosystem. With respect to technolo-gies, R&D activity can itself be thought of as aform of ‘search’ for new products and processes.However, R&D is too often usually a manifesta-tion of ‘local’ search. ‘Local’ search is only onecomponent of relevant search. In fast-paced envi-ronments, with a large percentage of new prod-uct introductions coming from external sources,search/exploration activity should not just be local.Enterprises must search the core as well as tothe periphery of their business ecosystem. Searchmust embrace potential collaborators—customers,suppliers, complementors—that are active in inno-vative activity.

Customers are sometimes amongst the first toperceive the potential for applying new technol-ogy. Visionary members of customer organizationsare often able to anticipate the potential for newtechnology and possibly even begin rudimentarydevelopment activities. Moreover, if the suppli-ers of new technology do not succeed in properlyunderstanding user/customer needs, it is unlikelythat new products they might develop will be suc-cessful. Indeed, one of the most consistent findingsfrom empirical research is that the probability thatan innovation will be successful commercially ishighly correlated with the developers’ understand-ing of user/customer needs (Freeman, 1974). Elec-tronic computing and the Internet itself can rightlybe viewed as having a significant component ofuser-led innovations. Business enterprises that arealert and sense the opportunity are often able toleverage customer-led efforts into new productsand services, as the users themselves are frequentlyill prepared to carry initial prototypes further for-ward.

Suppliers can also be drivers of innovationimportant in the final product. Innovation in micro-processor and DRAMs is a classic case. Thisupstream or ‘component’ innovation has impactedcompetition and competitive outcomes in personalcomputers, cellular telephony, and consumer elec-tronics more generally. Failure to ‘design in’ new

technology/components in a timely fashion willlend to failure; conversely, success can some-times be achieved by continuous rapid ‘design in.’Indeed, continuous and rapid design around newtechnology/components developed elsewhere canitself be a source of durable competitive advan-tage. Put differently, with rapid innovation by com-ponent suppliers, downstream competitive successcan flow from the ability of enterprises to continu-ously tap into such (external) innovation ahead ofthe competition. External search and acquisition oftechnology have been going on for decades, but asChesbrough (2003) explains, ‘Open Innovation’ isnow a mandate for enterprise success.

The concept and practice of open innovationunderscore the importance of broad-based externalsearch and subsequent integration involving cus-tomers, suppliers, and complementors. Establish-ing linkages between corporations and universitiesassists broad-based search, as university programsare usually unshackled from the near at hand.Indeed, a recent study of patenting in the opti-cal disk industry (Rosenkopf and Nerkar, 2001)seems to suggest that exploration that is more con-fined generates lower impacts, and that the impactof exploration is highest when exploration spansorganizational (but not technological) boundaries.However, it is not just a matter of searching forexternal inventions/innovations that represent newpossibilities. Frequently it is a matter of combiningcomplementary innovations so as to create a solu-tion to a customer problem. The systemic nature(Teece, 2000) of many innovations compounds theneed for external search.

Sensing opportunities and threats can also befacilitated if the enterprise and/or the entrepreneurexplicitly or implicitly employ some kind of ana-lytical framework, as this can help highlight whatis important. The field of strategic managementhas been stranded for some time with a frame-work that implicitly assumes that industry struc-ture (and product market share), mediated byenterprise behavior, determines enterprise perfor-mance. In Porter’s (1980) Five Forces frame-work, a good strategy involves somehow pickingan attractive industry and positioning oneself tobe shielded from competition. Porter’s approachmandates ‘industry’ analysis7 and the calibrationof five distinct industry-level forces: the role of

7 The Five Forces framework undergirds ‘industry’ analysis inbusiness school curriculum and in practice. However, the very



potential entrants, suppliers, buyers, substitutes,and rivalry amongst competitors. Because of itsrather static nature and the fact that it ignores manyaspects of the competitive environment includ-ing the role of complementarities, path dependen-cies, and supporting institutions, its application inthe contexts outlined in the Introduction of thispaper will limit the ability of the entrepreneurand/or the enterprise to properly sense opportu-nities and threats and properly calibrate strengths,weaknesses, and technological and market trajec-tories. If network effects, path dependencies, andthe co-evolution of technologies and institutionsare significant, the Five Forces framework is oflimited utility.

The Five Forces framework has inherent weak-nesses in dynamic environments. Fundamental isthat it implicitly views market structure as exoge-nous, when in fact market structure is the (endoge-nous) result of innovation and learning.8 Changesin science and technology create opportunities forinnovation. Enterprises can search amongst newpossibilities and engage in development activities.If successful, such development impacts the rela-tive fate of firms. This in turn determines marketstructure. Outcomes for individual enterprises areshaped in part by the selection processes at work inthe business ecosystem. Relevant factors ignoredor underplayed by Five Forces include techno-logical opportunities, path dependencies, appropri-ability conditions, supporting institutions, installedbase effects, learning, certain switching costs, andregulation. In short, in regimes of rapid technologi-cal change with well-developed markets for goodsand services (and poorly developed markets forknow-how), the Five Forces framework is com-promised because it has insufficient appreciation(a) for the importance of and nature of innova-tion and other factors that change the ‘rules of the

concept of an industry is itself of questionable value. If indus-try boundaries exist, they are faint, at least in technologicallyprogressive environments. For instance, the telecommunications‘industry’ may have had distinct boundaries over half a cen-tury ago around the telegraph and the telephone and associatedregulated services. However, by the 1960s, facsimile and dataservices had begun to be overlaid on the public telephone net-work. Today telephony is routinely carried by the Internet (usingVoice over IP) and cable TV networks.8 Indeed, the (basic) market structure–conduct–performanceparadigm from industrial economics that undergirds the FiveForces approach has been in need of revision for quite sometime. Phillips (1971) was perhaps the first to recognize that cau-sation is the reverse of what is assumed, with market structurebeing shaped by innovation.

game,’ (b) for factors inside the business enter-prise that constrain choices, (c) for factors thatimpact imitation and appropriability issues, (d) forthe role of supporting institutions, complementaryassets, cospecialization, and network externalities,or (e) for the blurred nature of industry boundaries.Also, as discussed later, in many ‘platform’ indus-tries or where there is significant outsourcing, scaleis an industry asset.

The dynamic capabilities framework representsa strong break with Five Forces. Within thedynamic capabilities framework, the ‘environmen-tal’ context recognized for analytical purposes isnot that of the industry, but that of the busi-ness ‘ecosystem’—the community of organiza-tions, institutions, and individuals that impact theenterprise and the enterprise’s customers and sup-plies. The relevant community therefore includescomplementors, suppliers, regulatory authorities,standard-setting bodies, the judiciary, and educa-tional and research institutions. It is a frameworkthat recognizes that innovation and its support-ing infrastructure have major impacts on com-petition. The dynamic capabilities framework isgrounded in Kirznerian, Schumpeterian, and evolu-tionary theories of economic change, whereas FiveForces is grounded in the Mason–Bain paradigmof industrial economics.9 Also, whereas accordingto Porter the essence of strategy formulation is‘coping with competition’ (Porter, 1991: 11), in thedynamic capabilities tradition the essence of strat-egy involves selecting and developing technolo-gies and business models that build competitiveadvantage through assembling and orchestratingdifficult-to-replicate assets, thereby shaping com-petition itself.

Even when utilizing the ecosystem as the orga-nizing paradigm for assessing developments in thebusiness environment, the full import of particularfacts, statistics, and developments is rarely obvi-ous. Accordingly, the evaluative and inferentialskill possessed by an organization and its manage-ment is important. Indeed, much of the informa-tion gathered and communicated inside the enter-prise has minimal decision relevance. Even if rel-evant, it often arrives too late. Management mustfind methods and procedures to peer through the

9 Developed in the 1930s, 1940s, and 1950s, it is still relevantto some of the ‘rust belt’ industries that experience low rates oftechnological innovation where complementors are not impor-tant, and where the coevolution of technologies and institutionsis not significant (Teece, 1990).


D. J. Teece

Processes to TapSupplier and

ComplementorInnovation.

Processes to Direct Internal R&D and Select

New Technologies.

Processes to IdentifyTarget Market Segments,

Changing CustomerNeeds, and Customer

Innovation.

Processes to TapDevelopments in

Exogenous Science andTechnology.

Analytical Systems (andIndividual Capacities) to

Learn and to Sense, Filter,Shape, and Calibrate

Opportunities.

Figure 1. Elements of an ecosystem framework for ‘sensing’ market and technological opportunities

fog of uncertainty and gain insight. This involvesgathering and filtering technological, market, andcompetitive information from both inside and out-side the enterprise, making sense of it, and figur-ing out implications for action. However, becauseattention is a scarce resource inside the enterprise(Cyert and March, 1963), management must care-fully allocate resources to search and discovery.The enterprise’s articulated strategy can becomea filter so that attention is not diverted to everyopportunity and threat that ‘successful’ searchreveals. Likewise, scenario planning can collapselikely situations into a small number of scenariosthat can facilitate cognition, and then action, onceuncertainty is resolved. Figure 1 summarizes indi-vidual and enterprise traits that undergird sensingcapabilities.

SEIZING OPPORTUNITIES

Nature of the capability

Once a new (technological or market) opportunityis sensed, it must be addressed through new prod-ucts, processes, or services. This almost alwaysrequires investments in development and commer-cialization activity. Multiple (competing) invest-ment paths are possible, at least early on. Thequintessential example is the automobile industry,where in the early days different engine technolo-gies—steam, electric, and gasoline—each hadtheir champions. Once a dominant design beginsto emerge, strategic choices become much morelimited. This paradigm, which was first offeredby Abernathy and Utterback (1978) and then builtupon by Teece (1986, 2007), now has considerableevidence supporting it over a wide range of tech-nologies (Klepper and Graddy, 1990; Utterback

and Suarez, 1993; Malerba and Orsenigo, 1996).It implicitly recognizes inflexion points in tech-nological and market evolution. These inflexionpoints impact investment requirements and strate-gic choices. Implications for investment decisionshave been noted elsewhere (Teece, 1986) andinclude staying flexible until the dominant designemerges and then investing heavily once a designlooks like it can become the winner. Any strat-egy is, of course, likely to be fraught with hazardsbecause of uncertainties. Moreover, the mannerand time at which an enterprise needs to place itsbets depend on competition in the ‘input’ marketsand on the identity of the enterprise itself. Mitchell(1991) suggests that the timing of resource com-mitments can differ according to the enterprise’sexisting positions with respect to the relevant com-plementary assets. Enterprises that are well posi-tioned can wait, while those that are not mustscramble.

Addressing opportunities involves maintainingand improving technological competences andcomplementary assets and then, when the opportu-nity is ripe, investing heavily in the particular tech-nologies and designs most likely to achieve mar-ketplace acceptance. When network externalitiesare present, early entry and commitment are nec-essary. The presence of increasing returns meansthat if one network gets ahead, it tends to stayahead. Getting ahead may require significant up-front investments. Customers will not want anenterprise’s products if there are strong networkeffects and the installed base of users is rela-tively small. Accordingly, one needs to strategizearound investment decisions, getting the timingright, building on increasing return advantages, andleveraging products and services from one applica-tion to another. The capacity to make high-quality,



unbiased but interrelated investment decisions inthe context of network externalities, innovation,and change is as rare as decision-making errorsand biases are ubiquitous.

However, the issue that the enterprise faces isnot just when, where, and how much to invest.The enterprise must also select or create a partic-ular business model that defines its commercial-ization strategy and investment priorities. Indeed,there is considerable evidence that business suc-cess depends as much on organizational innova-tion, e.g., design of business models, as it does onthe selection of physical technology. This is trueat the enterprise level as well as at the economy-wide level (Nelson, 2005). Indeed, the inventionand implementation of business models and asso-ciated enterprise boundary choices involve issuesas fundamental to business success as the devel-opment and adoption of the physical technologiesthemselves. Business models implicate processesand incentives; their alignment with the physicaltechnology is a much overlooked component ofstrategic management. The understanding of theinstitutional/organizational design issues is typi-cally more limited than the understanding of thetechnologies themselves. This ignorance affordsconsiderable scope for mistakes around the properdesign of business models and the institutionalstructures needed to support innovation in both theprivate and public sectors.

In theory, one could imagine transactionsbetween entities that scout out and/or developopportunities, and those that endeavor to executeupon them. In reality, the two functions cannotbe cleanly separated, and the activities must beintegrated inside a single enterprise, where newinsights about markets—particularly those thatchallenge the conventional wisdom—will likelyencounter negative responses. The promoters/visionaries must somehow defeat the naysayers,transform internal views, and facilitate necessaryinvestment. Some level of managerial consensuswill be necessary to allow investment decisions tobe made. Investment will likely involve commit-ting financial resources behind an informed con-jecture about the technological and marketplacefuture. However, managers of established productlines in large organizations can sometimes havesufficient decision-making authority to starve thenew business of financial capital. This posture canbe buttressed by capital budgeting techniques thatmore comfortably support investments for which

future cash flow can be confidently projected. Inshort, the new can lose out to the established unlessmanagement is sensitive to the presence of cer-tain biases in accepted investment decision pro-cesses. An important class of dynamic capabilitiesemerges around a manager’s ability to overridecertain ‘dysfunctional’ features of established deci-sion rules and resource allocation processes.

It helps to begin by recognizing that decision-making processes in hierarchically organized enter-prises involve bureaucratic features that are use-ful for many purposes, but they nevertheless maymuzzle innovation proclivities. In particular, a for-mal expenditure process involving submissionsand approvals is characteristic of ‘well-managed’companies. Decision making is likely to have acommittee structure, with top management requir-ing reports and written justifications for signif-icant decisions. Moreover, approvals may needto be sought from outside the organizational unitin which the expenditure is to take place. Whilethis may ensure a matching up of expendituresto opportunities across a wider range of economicactivity, it unquestionably slows decision makingand tends to reinforce the status quo. Commit-tee decision-making structures almost always tendtoward balancing and compromise. But innovationis often ill served by such structures, as the newand the radical will almost always appear threat-ening to some constituents. Strong leaders can fre-quently overcome such tendencies, but such lead-ers are not always present. One consequence is a‘program persistence bias.’ Its corollary is variousforms of ‘anti-innovation bias,’ including the ‘anti-cannibalization’ basis discussed in a later section.Program persistence refers to the funding of pro-grams beyond what can be sustained on the merits,and follows from the presence or influence of pro-gram advocates in the resource allocation process.This proclivity almost automatically has the coun-tervailing effect of reducing funds available to newinitiatives.

One should not be surprised, therefore, if anenterprise senses a business opportunity but failsto invest. In particular, incumbent enterprises tendto eschew radical competency-destroying innova-tion in favor of more incremental competency-enhancing improvements. The existence of layerupon layer of standard procedures, establishedcapabilities, complementary assets, and/or admin-istrative routines can exacerbate decision-makingbiases against innovation. Incumbent enterprises,


D. J. Teece

relying on (path-dependent) routines, assets, andstrategies developed to cope with existing tech-nologies, are handicapped in making and/or adopt-ing radical, competency-destroying, noncumula-tive innovation (Nelson and Winter, 1982; Tush-man and Anderson, 1986; Henderson and Clark,1990). This is true whether the competence isexternal to the firm or internal to the firm.

Evidence also shows that decision-makers dis-count outcomes that are merely probable in com-parison with outcomes that are certain. This hasbeen called the certainty effect (Kahneman andLovallo, 1993). It contributes to excessive riskaversion when choices involve possible losses.Further, to simplify choices between alternatives,individuals generally evaluate options in isolation.Viewing each alternative as unique leads decision-makers to undervalue possibilities for risk pooling.This approach to decision making may produceinconsistent preferences and decision biases (timidchoices) that lead to outcomes that block inno-vation (Kahneman and Tversky, 1979; Kahnemanand Lovallo, 1993). An opposing bias to loss/riskaversion is excessive optimism. This leads toinvestment in low or negative return projects. As aresult, entry decisions often fail. Audretsch (1995)found that over the period 1976–86 the average10-year failure rate in two-digit SIC manufacturingsectors ranged from 75.8 percent to 54.8 percent.Similar failure rates have been reported in otherstudies (Dunne, Roberts, and Samuelson, 1988;Klepper and Miller, 1995). However, these failurerates disguise wide variation amongst particularenterprises and between new entrants and incum-bents.

The existence of established assets and routinesexacerbates problems of excessive risk aversion.Specifically, both the isolation effect and the cer-tainty effect can be intensified by the existence ofestablished assets, causing incumbent enterprisesto become comparably more risk averse than newentrants. In terms of innovative activity, this exces-sive risk aversion leads to biased decision makingand limits the probability that incumbent enter-prises will explore risky radical innovations. Inshort, success in one period leads to the establish-ment of ‘valid’ processes, procedures, and incen-tives to manage the existing business. This canhave the unintended effect of handicapping thenew business. The proficiency with which suchbiases are overcome and a new opportunity isembraced is likely to depend importantly on the

quality of the enterprise’s routines, decision rules,strategies, and leadership around evaluating newinvestment opportunities. Business historians (e.g.,Chandler, 1990a; Lazonick, 2005) and others havereminded us that over the long run the ability ofenterprises to commit financing and invest astutelyaround new technologies is critical to enterpriseperformance.10

In regimes of rapid technological innovation, itis clear that making investment choices requiresspecial skills not ubiquitously distributed amongstmanagement teams. Nor are they ubiquitously dis-tributed amongst investors.11 Resource/asset align-ment and coalignment issues are important inthe context of innovation, but they are quitedifferent from portfolio balance issues faced byfinancial investors. The presence of increasingreturns means that one also needs to strategizearound investment decisions, getting the timingright, building on increasing return advantages, andleveraging products and services from one appli-cation to another. Value-enhancing investmentsinside the knowledge-based enterprise are oftencospecialized12 to each other. Also, the nature ofthe portfolio ‘balance’ needed inside the enterpriseis different from the portfolio balance sought bypure financial investors. The economics of cospe-cialization are not the economics of covariancewith which investors are familiar. In short, thetask of making astute project- and enterprise-levelinvestment decisions is quite challenging becauseof cospecialization, and irreversibilities.

The project finance and related literatures pro-vide tools and clear decision rules for projectselection once cash flows are specified, uncertaintyand/or risk are calibrated, and interdependencies

10 Consider the development of civilian jet transport aircraft inthe United States in the 1950s. As Phillips (1971: 126) noted:‘Any one of Boeing, Douglas, Lockheed, or Corvair mighthave been first. . . . The technology was there to adapt to—notrisklessly or costlessly to be sure, but it was there. Perhaps thebiggest risk in 1953 was not technological in character. Instead,it was risk with respect to what sort of jet to build and when tobuild it.’11 The decision skills required of management have limitedcommonality with those of an investor. One difference is theilliquidity and irreversibility of most managerial investmentdecisions. Another is the need to achieve continuous alignmentamongst the assets at work in the enterprise. Both public andprivate equity investors typically lack this kind of orchestrationand integration capability or capacity. Moreover, their skills aremost applicable when investments are liquid.12 Cospecialization is defined and discussed in Teece (1986) andexplored further in the section below entitled ‘Managing threatsand reconfiguration.’



between and amongst cash flows are ignored.However, the essence of the investment decisionfor the (strategic) manager is that it involves esti-mating interdependent future revenue streams andcost trajectories, and understanding a panoply ofcontinuous and interrelated cospecialized invest-ment issues.13 The returns to particular cospecial-ized assets cannot generally be neatly apportionedor partitioned. As a result, the utility of traditionalinvestment criteria is impaired. Thus while project-financing criteria (e.g., discounted cash flow, pay-back periods and the like) and techniques for deci-sion making under uncertainty are well known,there is little recognition of how to value intangi-bles and take into account features such as cospe-cialization, irreversibility, and opportunity costs.14

Nor is the concept of a ‘strategic investment’ rec-ognized in the finance literature. Finance theoryprovides almost no guidance with respect to how toestimate future cash flows, although making suchestimates is as much, if not more, the essence ofgood decision making as are the methodologiesand procedures for analyzing cash flow.

In short, managers need to make unbiased judg-ments under uncertainty around not just futuredemand and competitive responses associated withmultiple growth trajectories, but also around thepay-offs from making interrelated investments inintangible assets. In the world of tangible assets,this can sometimes be precisely modeled; notso for the world of cospecialized intangibles.In essence, the organizational challenge appearsto be that in environments experiencing rapidchange, activities are not fully decomposable.Cross-functional activities and associated invest-ments must take place concurrently, rather thansequentially, if enterprises are to cut time-to-market for new products and processes. Man-agerial judgments (decision-making skills) takeon great significance in such contexts. This wasalso true during prior centuries, as Alfred Chan-dler’s (1990a, 1990b) analysis of successful enter-prises from the 1870s through the 1960s makes

13 Monteverde and Teece’s (1982) study of the automobile indus-try showed that ‘systems integration’ considerations impactedmake–buy decisions. This evidence hints at the value to becreated from figuring out heuristics and protocols likely to aiddecisions involving interrelated investments. Evans, Hagiu, andSchmalensee (2006) recognize multisided market interdependen-cies which likewise require a systems perspective.14 Ghemawat (1991) and many others have examined uncertaintyand irreversibilities. However, cospecialization has received verylittle attention.

apparent. No matter how much analytical work isdone, tacit investment skills are of great impor-tance. Chandler further argues that success in thelate-nineteenth and much of the twentieth centurycame to those enterprises that pursued his ‘three-pronged’ strategy: (1) early and large-scale invest-ments behind new technologies; (2) investment inproduct-specific marketing, distribution, and pur-chasing networks; and (3) recruiting and organiz-ing the managers needed to supervise and coordi-nate functional activities. The first and second ele-ments require commitment to investments whereirreversibilities and cospecialization are identified.While the nature of required investments mayhave changed in recent decades (less decompos-able/more interrelated), investment decision skillsremain important.

Microfoundations

Selecting product architectures and businessmodels

The design and performance specification of prod-ucts, and the business model employed, all helpdefine the manner by which the enterprise deliv-ers value to customers, entices customers to payfor value, and converts those payments to profit.They reflect management’s hypothesis about whatcustomers want and how an enterprise can bestmeet those needs, and get paid for doing so. Theyembrace: (1) which technologies and features areto be embedded in the product and service; (2) howthe revenue and cost structure of a business is to be‘designed’ and if necessary ‘redesigned’ to meetcustomer needs; (3) the way in which technolo-gies are to be assembled; (4) the identity of marketsegments to be targeted; and (5) the mechanismsand manner by which value is to be captured.The function of a business model is to ‘articu-late’ the value proposition, select the appropriatetechnologies and features, identify targeted marketsegments, define the structure of the value chain,and estimate the cost structure and profit potential(Chesbrough and Rosenbloom, 2002: 533–534). Inshort, a business model is a plan for the organi-zational and financial ‘architecture’ of a business.This model makes assumptions about the behav-ior of revenues and costs, and likely customer andcompetitor behavior. It outlines the contours of thesolution required to earn a profit, if a profit is avail-able to be earned. Once adopted it defines the waythe enterprise ‘goes to market.’ Success requires


D. J. Teece

that business models be astutely crafted. Other-wise, technological innovation will not result incommercial success for the innovating enterprise.Generally there is a plethora of business modelsthat can be designed and employed, but some willbe better adapted to the ecosystem than others.Selecting, adjusting, and/or improving the businessmodel is a complex art.

Nevertheless, the importance of ‘business mod-els’ has been given short shrift in the academicliterature, at least until quite recently. Important(business model) choices include technologicalchoices, market segments to be targeted, financialterms (e.g., sales vs. leasing), choices with respectto bundled vs. unbundled sales strategies, jointventures vs. licensing vs. go-it-alone approaches,etc. For example, in the early days of the copierindustry, Xerox focused on leasing rather thanselling copiers. This stemmed from a belief thatcustomer trial would lead to further use. Anotherexample from the United States is Southwest Air-lines, which believes that most customers wantlow frills, reliability, and low cost. It eschewsthe hub-and-spoke model, does not belong to anyalliances, and does not allow interlining of passen-gers and baggage. Nor does it sell tickets throughtravel agencies—all sales are direct. All aircraftare Boeing 737s. Its business model is quite dis-tinct from the major carriers, although many havetried (without much success) to copy elements ofthe Southwest model.15

The capacity an enterprise has to create, adjust,hone, and, if necessary, replace business modelsis foundational to dynamic capabilities. Choicesaround how to capture value all help determinethe architecture or design of a business. Havinga differentiated (and hard-to-imitate) yet effectiveand efficient ‘strategic architecture’ to an enter-prise’s business model is important. Both Dell Inc.and Wal-Mart have demonstrated the value associ-ated with their business models (Webvan and many

15 Let us take another example. A rock star might decide to useconcerts as the key revenue generator, or the concert may beused primarily to stimulate sales of recordings. The star coulddecide to spend less time performing at concerts, and more timein the recording studio. There is clearly a choice of variousmedia to extract value: live productions, movies, sale of CDsthrough stores, online sale of music through virtual stores suchas the iTunes store offered by Apple, etc. The emergence ofthe Internet, Napster, and Napster clones in turn requires artists(and record companies) to rethink their business models. Theability to reconfigure business models for delivering and pricingmusic profitably is undoubtedly a dynamic capability for boththe record companies and the artists.

other dot-coms demonstrated just the opposite).Both Dell Inc.’s and Wal-Mart’s business modelswere different, superior, and hard for competitorsto replicate. They have also constantly adjustedand improved their processes over time.16

One might be tempted to argue that designing,implementing, and validating business models isstraightforward, but this simply is not so. Aspectsof designing (and redesigning) a business modelare undoubtedly readily routinized and codified.Note the plethora of business books providinginstruction on how to write a business plan. Suchmanuals can provide some discipline to the busi-ness model questions that one should ask. How-ever, designing a new business requires creativity,insight, and a good deal of customer, competitor,and supplier information and intelligence. There isa significant tacit component. Entrepreneurs andexecutives are forced to make many informedguesses about customer and competitor behavior,as well as the behavior of costs. Indeed, validat-ing a business model and a business plan requiresboth effort and judgment. It takes detailed fact-specific inquiry including: a keen understanding ofcustomer needs and customer willingness to pay;an understanding of procurement cycles and thesales cycle; knowledge of supply and distributioncosts; and an understanding of competitor position-ing and likely competitive responses. Put differ-ently, selecting the right ‘architecture’ for a busi-ness requires not just understanding the choicesavailable; it also requires assembling the evidenceneeded to validate conjectures and hunches aboutcosts, customers, competitors, complementors, dis-tributors, and suppliers.

Designing good business models is in part‘art.’ However, the chances of success are greaterif enterprises (1) analyze multiple alternatives,(2) have a deep understanding of user needs,(3) analyze the value chain thoroughly so as tounderstand just how to deliver what the customerwants in a cost-effective and timely fashion, and(4) adopt a neutrality or relative efficiency per-spective to outsourcing decisions. Useful toolsinclude market research and transaction cost eco-nomics. Chesbrough and Rosenbloom (2002) sug-gest that established enterprises often have blinders

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



with respect to alternative business models—andthat this prevails even if the technology is spunoff into a separate organization, where other (path-dependent constraints) are less likely to exist.

In short, designing the business correctly, andfiguring out what John Seeley Brown refers toas the ‘architecture of the revenues’17 (and costs),involve processes critical to the formation and suc-cess of new and existing businesses. No amount ofgood governance and leadership is likely to leadto success if the wrong business model is in place.Good business models achieve advantageous coststructures and generate value propositions accept-able to customers. They will enable innovators tocapture a large enough portion of the (social) valuegenerated by innovation18 to permit the enterpriseat least to earn its cost of capital.

Selecting enterprise boundaries

In regimes of rapid technological progress, settingthe enterprise boundaries correctly is important,and can be viewed as an element of getting thebusiness model right. In Teece (1986), Chesbroughand Teece (1996), and Teece (1986, 2007) norma-tive rules were advanced indicating how enterpriseboundaries ought to be set to ensure that innovationis more likely to benefit the sponsor of the inno-vation rather than imitators and emulators. Keyelements of this framework were: (1) the appropri-ability regime (i.e., the amount of natural and legalprotection afforded the innovation by the circum-stances prevailing in the market); (2) the natureof the complementary assets (cospecialized or oth-erwise) that an innovating enterprise possessed;(3) the relative positioning of innovator and poten-tial imitators with respect to complementary assets;and (4) the phase of industry development (pre orpost the emergence of a dominate design). Theframework is prescriptive not only as to strategybut also as to outcomes.

Enterprise boundary decisions need to reflectother criteria too. A company’s integration up-

17 Quoted in Chesbrough and Rosenbloom (2002: 529).18 A recent effort to establish a new business model is exem-plified by the efforts of Rambus to rely exclusively on patentlicensing to capture value from its significant technological con-tributions to the design of semiconductor memory devices. Suchan approach avoids building fabrication facilities (which areextremely expensive) but its viability depends entirely on Ram-bus’s ability to enforce its patents in an environment in whichcourts are sometimes reluctant to enjoin infringers and whereenforcing broad patents may engender antitrust challenges.

stream, downstream, as well as externally, ispartly driven by the need to build capabilities,particularly when such capabilities are not widelydistributed in the industry. Of course, vertical spe-cialization is not itself independent of enterprisestrategy, and vice versa (Macher and Mowery,2004). Studies of the early vertical evolution ofthe petroleum industry stressed the need to alignupstream and downstream capacities in an envi-ronment where qualified business partners werelimited (Teece, 1976). Pisano, Shan, and Teece(1988: 202) developed a framework for under-standing R&D outsourcing that recognized that thelocus of world-class research/productive capabil-ity might lie external to the enterprise, requiringoutsourcing as a way to compete.19 Jacobides andWinter (2005: 398) have also clearly stated that ‘itis necessary to look at the distribution of produc-tive capabilities—to understand when enterprisesare integrated and when they are not. It becomesclear that vertical specialization must be in part afunction of heterogeneity in productive capabili-ties along the value chain.’ They also note that thecapability development process itself changes as aconsequence of changing scope. Recognition thatsystemic innovation favors integration, for bothtransaction costs and capability reasons, is alsoembedded in the saga of the development of thediesel electric locomotive (Teece, 1988). The abil-ity of enterprises to procure technology externallyas well as develop it internally are critical skills,as discussed above and in Teece (1986), Ches-brough and Teece (1996), and Teece (2000). Firmsmust dispel prejudices against technology fromthe outside, and hone their absorptive capacitythrough learning activities and skill accumulation.Enterprises may require alliance arrangements toactively learn and upgrade relevant skills (Branzeiand Vertinsky, 2006).

The critical strategic element associated withcapturing value from innovation is the ability ofthe innovating enterprise to identify and control the‘bottleneck assets’ or ‘choke points’ in the valuechain from invention through to market (Teece,1986, 2006). Outsourcing those assets/services that

19 The model identified transaction costs, the locus of capabilities(inside or outside the enterprise), and appropriability regimesas three relevant classes of factors driving enterprise boundarydecisions. In particular, it was noted that transaction cost factors‘must be weighed against any losses in productive efficiencythat result from being less skilled than specialists in the relevantstages of production.’


D. J. Teece

are in competitive supply is, of course, consis-tent with such a strategy. In short, the boundariesof the enterprise need to be artfully contouredfor each major innovation, using decision crite-ria referenced above. Failure to do so is likely tobe associated with the failure to stimulate marketdevelopment (especially of complementary tech-nologies) and incomplete capture of the profitsavailable from innovation.

Managing complements and ‘platforms’

Investment choices in many high-technologyindustries today are driven by imperatives quitedifferent from the (industrial) contexts that haveanimated strategy research over the past half cen-tury. Scale and scope economy ‘mandates,’ whichto some strategists dictate the scale and scope ofthe enterprise, have given way to a different setof mandates around developing (or encouraging)complementary investments and capturing cospe-cialization benefits. The reason for this is that inmany industries outsourcing has made scale anindustry asset, in the sense that economies of scalecan be captured by outsourcing to contract manu-facturers who, in the face of competition, pass onthe benefits of scale. Witness the contract semicon-ductor fabricators. They enable fabless semicon-ductor ‘designers’ to capture most of the benefitsof scale without engaging in manufacturing. Like-wise, in the clothing industry, small-scale design-ers of footwear and outerwear can source at com-petitive rates from large suppliers, thereby cap-turing the benefit of scale economics previouslyenjoyed only by large integrated manufacturers.With competition, scale advantages are not propri-etary, and are unlikely to be a source of sustainabledifferentiation.

When intermediate (product) markets are welldeveloped, neither economies of scale noreconomies of scope need define the scale and scopeof the enterprise. Contractual access (on compet-itive terms) to scale-based ‘facilities’ vitiates theneed for enterprise scale and scope. This was themajor theme in Teece (1980) but the importanceof the argument was often not appreciated. Todayits importance is more evident.

While the importance of scale and scopeeconomies to enterprise boundary decisions mayhave been softened, the significance to enter-prise strategy of cospecialization has been ele-vated. As viewed by customers, high-technology

‘products’ are often systems. These systems con-sist of interdependent components resting on ‘plat-forms.’ There is strong functional interdependenceamongst components of the system. End userdemand is for the system, not the platform. Thereis often a multisided ‘market’ phenomenon at workas well. For instance, electronic game consoles arenot much use without games; computer operatingsystems are not much use without a suite of appli-cation programs; credit cards are not much useto cardholders without merchants that will acceptthem, and vice versa; and hydrogen cars are notmuch use without hydrogen filling stations, andvice versa. This important class of situations hashighlighted the importance of cospecialization, andstrategic decision making must now take this intoaccount.

The phenomenon is not new—the automobileindustry depended first on the general store andthen specialized retail outlets to make gasolineubiquitously available to motorists. The role ofcomplementary assets and cospecialization hasalready been recognized in the innovation process,and a decision framework outlined to chart theinnovator on a course more likely to lead to ahigher share of the available profit (Teece, 1986,2007). What is new is that complements often siton top of what might be thought of as ‘platforms,’which are managed by an incumbent enterprise(Evans et al., 2006). In these circumstances, entrydecision and ‘boundary’ conundrums exist. Theplatform owner needs complementary products tobe provided by others, particularly when it haslittle or no relevant skills to develop them itself.Fostering innovation and entry by the providers ofcomplementary products may, in fact, require theplatform manager to commit (by word or deed) notto provide certain complements. When the inter-face between the complementors and the platformis itself evolving, decision rules become ever morecomplex. The platform owner and the complemen-tors might also need to consider whether the plat-form needs to be open or proprietary, and whethertools and other incentives should be provided tostimulate investment by the complementors. Deci-sion frameworks that recognize the importanceof network effects, dispersion in the sources ofinnovation of complementary products, interoper-ability issues, and installed base trajectories mustall be factored into decisions. Quality decisionswill require uncommon foresight and the ability toshape outcomes. In this regard, the existing asset



base of the platform manager, including its finan-cial resources, is of considerable significance. Thedistribution of (development) capabilities betweenthe platform manager and the complementors willalso be important. Also, as discussed below, theboundaries of the enterprise (i.e., whether the plat-form manager is also providing complements) islikely to be of significance, possibly deterring (orencouraging) entry and innovation by complemen-tors.

Avoiding bias, delusion, deception, and hubris20

As noted, proclivities toward decision errors arenot uncommon in managerial decision making,particularly in large organizations. Investment deci-sion errors already identified include excessiveoptimism, loss aversion, isolation errors, strate-gic deception, and program persistence. As Nelsonand Winter (2002: 29) note, organizational deci-sion processes often display features that seem todefy basic principles of rationality and sometimesborder on the bizarre. These errors can be espe-cially damaging in fast-paced environments withpath dependencies and network effects, as there isless opportunity to recover from mistakes. Wheninvestments are small and made frequently, thereare many opportunities to learn from mistakes.Since large investments are usually occasional,major investment decisions are likely to be (poten-tially) more vulnerable to error.

Fortunately, biases can be recognized ahead oftime. Enterprises can bring discipline to bear topurge bias, delusion, deception, and hubris. How-ever, the development of disciplines to do so is stillin its infancy. The implementation of proceduresto overcome decision-making biases in enterprisesettings is, accordingly, not yet a well-distributedskill, and may not be for decades to come. Accord-ingly, competitive advantage can be gained byearly adopters of techniques to overcome decisionbiases and errors.

Overcoming biases almost always requires acognitively sophisticated and disciplined approachto decision making. Being alert to the incentivesof the decision-makers and to possible informa-tion asymmetries is a case in point. Obtaining an‘outside view’ through the review of external datacan help eliminate bias. Testing for errors in logic

20 I would like to thank Dan Lovallo for inspiration and help inthis section.

is also essential. Management also needs to createan environment where the individuals involved inmaking the decision, at both the management andboard level, feel free to offer their honest opinions,and look at objective (historical) data in order toescape from closed thinking. Incentives must alsobe designed to create neutrality when assessinginvestments in the old and the new.

Considerable progress in combating biases hasbeen made. Advisors call upon managers to adoptradical, nonformulaic strategies in order to over-come the inertias that inhibit breakthrough inno-vation (Davidow and Malone, 1992; Handy, 1990).Specifically, corrective strategies encouragechange through two basic mechanisms: (1) design-ing organizational structures, incentives and rou-tines, to catalyze and reward creative action; and(2) developing routines to enable the continualshedding of established assets and routines that nolonger yield value. Strategies that provide struc-tures, incentives, and processes to catalyze andreward creative action serve to attenuate problemsof excessive risk aversion. For example, strategiesthat call on the enterprise to ‘cut overhead’ and‘increase divisional authority’ can be interpreted asefforts to reduce the number of management lay-ers of the enterprise and to push decision makingdown to lower levels to minimize the inherent iso-lation errors associated with multilevel, hierarchi-cal decision-making processes. These recommen-dations can be viewed as organizational processesand strategic mechanisms to mitigate decision-making biases.

Perhaps most importantly, executives mustacknowledge the interaction effect between own-ing established assets and decision-making biases.Many recommended strategies (such as cannibal-izing profitable product lines and licensing yourmost advanced technology) call for the shed-ding of established capabilities, complementaryassets and/or administrative routines to reduce theintensity of decision-making biases. By jettisoning‘dead’ or dying assets, the enterprise is no longershackled with an asset base that can be a crutchand provide a false sense of security, and sustaingroups inside the enterprise that persist in torpedo-ing new initiatives. In abandoning dead or dyingassets, the enterprise frees itself of certain routines,constraints, and opportunities for undesirable pro-tective action inside the enterprise.

Sources of the ‘anti-cannibalization’ bias men-tioned earlier can also be attacked. Self-serving


D. J. Teece

behavior inside the enterprise to ‘protect’ incum-bent constituencies undergirds this bias. Flawedinvestment frameworks may also contribute. Entryinto a market by an enterprise with a new andsuperior technology will cause rapid deprecia-tion of the economic value of an incumbent’splant and equipment. However, the incumbent maywell make business decisions based on exam-ining accounting profits that reflect depreciationrates specified by accepted accounting standards.If decision-makers confuse depreciation calculatedaccording to generally accepted accounting prin-ciples (GAAP) with real economic depreciation,and conclude that the existing business is stillprofitable when, in fact, it is not, then the busi-ness enterprise may eschew profit-enhancing can-nibalization of its own products. To guard againstthis bias, investment decision-makers and incum-bents must use accounting data cautiously. In par-ticular, they must also consider the opportunitycost associated with not cannibalizing their ownproducts. Capital-budgeting procedures implicitlybiased against projects with long-term horizonsmust be jettisoned or used cautiously. That is notto say that incumbents need to invest on the sameschedule as new entrants. As Teece (1986) andMitchell (1991) demonstrate, incumbents need notbe the first movers. Superior positioning in com-plementary assets may enable incumbents to letthe new entrants do the prospecting, investing lateronce market and technological risk has diminished.

There is an obvious role for leadership in makingquality decisions, communicating goals, values,and expectations, while also motivating employees

and other constituencies. Organizational identifi-cation (and commitment, which is the corollary)can dramatically augment enterprise performance,although it is doubtful it can override completelymisaligned incentives. Nevertheless, group loy-alty is a ‘powerful altruistic force’ that conditionsemployee goals and the cognitive models theyform of their situation (Simon, 1993: 160). Topmanagement through its action and its commu-nication has a critical role to play in garneringloyalty and commitment and achieving adherenceto innovation and efficiency as important goals.Since there is already an extensive literature onculture, commitment, and leadership, these issuesare not discussed further. However, it would bea significant oversight in a summary statement ofthe dynamic capabilities framework to ignore themcompletely. Their full integration into the frame-work is left to others. However, it is recognizedthat to the extent such properties are not ubiq-uitously distributed amongst business enterprises,they can be a very important source of superiorperformance. Figure 2 summarizes the microfoun-dations identified in this section of the paper.

MANAGING THREATS ANDRECONFIGURATION

Nature

The successful identification and calibration oftechnological and market opportunities, the judi-cious selection of technologies and product attri-butes, the design of business models, and the

Selecting Decision-Making Protocols

Recognizing Inflexion Points andComplementarities; Avoiding Decision Errors andAnticannibalization Proclivities.

Enterprise Structures,Procedures, Designs and

Incentives for SeizingOpportunities

Selecting the Technology and ProductArchitecture;Designing Revenue Architectures;Selecting Target Customers;Designing Mechanisms to CaptureValue.

Delineating the Customer Solution andthe Business Model

Demonstrating Leadership; Effectively Communicating;Recognizing Non-Economic Factors,Values, and Culture.

Building Loyalty and Commitment

Calibrating Asset Specificity;Controlling Bottleneck Assets;Assessing Appropriability; Recognizing, Managing, andCapturing CospecializationEconomies.

Selecting Enterprise Boundaries toManage Complements and “Control”

Platforms

Figure 2. Strategic decision skills/execution



commitment of (financial) resources to investmentopportunities can lead to enterprise growth andprofitability. Profitable growth will lead to the aug-mentation of enterprise-level resources and assets.Success will cause the enterprise to evolve in apath-dependent way. A key to sustained profitablegrowth is the ability to recombine and to recon-figure assets and organizational structures as theenterprise grows, and as markets and technolo-gies change, as they surely will. Reconfigurationis needed to maintain evolutionary fitness and, ifnecessary, to try and escape from unfavorable pathdependencies. In short, success will breed somelevel of routine, as this is necessary for operationalefficiency. Routines help sustain continuity untilthere is a shift in the environment. Changing rou-tines is costly, so change will not be (and shouldnot be) embraced instantaneously. Departure fromroutines will lead to heightened anxiety within theorganization, unless the culture is shaped to accepthigh levels of internal change. If innovation isincremental, routines and structures can probablybe adapted gradually or in (semi-continuous) steps.When it is radical, possibly because it is sciencebased, then there will be a mandate to completelyrevamp the organization and create an entirely new‘break out’ structure (Teece, 2000) within which anentirely different set of structures and proceduresis established.

As discussed earlier, the ‘anti-cannibalization’bias is a particular manifestation of incentiveand structural problems that can thwart innova-tion in established enterprises. Incumbent enter-prises possessing fixed assets may further tend tolimit their new investments to innovations thatare ‘close-in’ to the existing asset base. Theytend to narrowly focus search activities to exploitestablished technological and organizational assets.This effect makes it difficult for these enterprisesto see potential radical innovations. In addition,incumbent enterprises tend to frame new prob-lems in a manner consistent with the enterprise’scurrent knowledge base, assets, and/or establishedproblem-solving heuristics and established busi-ness model. This second effect means that man-agers may not successfully address opportunities orpotential innovations even when they do recognizethem. Managers face and must overcome at leasttwo constraints—cognitive limitations and fram-ing biases—arising from established assets (Teece,2000).

As the enterprise grows, it has more assets tomanage and to protect against malfeasance andmismanagement. Shirking, free riding, the strategicmanipulation of information, and internal compla-cency are all issues that established enterpriseswill confront continuously. As discussed earlier,over time successful enterprises will develop hier-archies and rules and procedures (routines) thatbegin to constrain certain interactions and behav-iors unnecessarily. Except in very stable envi-ronments, such rules and procedures are likelyto require constant revamping if superior perfor-mance is to be sustained. It is not uncommon tofind that a once functional routine becomes dys-functional, providing inertia and other rigiditiesthat stand in the way of improved performance(Leonard-Barton, 1995; Rumelt, 1995). As a result,less well-resourced enterprises (sometimes estab-lished enterprises that have divested certain assets,sometimes new entrants) end up winning in themarketplace.

Traditional management approaches endorsestrong hierarchies with at least three levels ofmanagement: top, middle, and lower. Control isexerted at the top and cascades down through mul-tiple levels. Employees tend to end up beholdento the management and CEO, and not the cus-tomer. The existence of independent profit centerscan lead to internal boundaries that stand in theway of providing integrated solutions that bene-fit customers. With centralized structures, strategicdecisions made at the top tend to become iso-lated from marketplace realities. Customer care isrelegated to employees who are lower down inthe organization. In short, the systems and rulesneeded to manage many layers of organizationtend to create structural rigidities and perversitiesthat in turn handicap customer and technologicalresponsiveness. To sustain dynamic capabilities,decentralization must be favored because it bringstop management closer to new technologies, thecustomer, and the market.

Top management leadership skills are requiredto sustain dynamic capabilities. An important man-agerial function is achieving semi-continuous assetorchestration and corporate renewal, including theredesign of routines. This is because the sustainedachievement of superior profitability requires semi-continuous and/or continuous efforts to build,maintain, and adjust the complementarity of prod-uct offerings, systems, routines, and structures.Inside the enterprise, the old and the new must


D. J. Teece

complement. If they do not, business units mustbe disposed of or placed in some type of sepa-rate structure. Otherwise, work will not proceedefficiently, and conflicts of one kind or anotherwill arise. Put differently, periodic if not continu-ous asset orchestration—involving achieving assetalignment, coalignment, realignment, and rede-ployment—is necessary to minimize internal con-flict and to maximize complementarities and pro-ductive exchange inside the enterprise.

Redeployment and reconfiguration (Capron,Dussauge, and Mitchell, 1998) may also involvebusiness model redesign as well as asset-realign-ment activities, and the revamping of routines.Redeployment can involve transfer of nontrad-able assets to another organizational or geographiclocation (Teece, 1977, 1980). It may or may notinvolve mergers, acquisitions, and divestments.21

Helfat and Peteraf (2003: 1006) suggest that capa-bility redeployment takes one of two forms: thesharing of capability between the old and the new,and the geographic transfer of capability from onemarket to another. Both are possible, but neither iseasy.

Microfoundations

Achieving decentralization and neardecomposability

Every system comprises subsystems (elements)that are to some extent interdependent and inde-pendent. However, as discussed earlier, enterprisesare unlikely to be continuously responsive to cus-tomers and new technologies absent a high degreeof decentralization. With decentralized decisionmaking, different managers observe different infor-mation and control different decisions, but there isnot the need for communication to a single centraldecision-maker, and hence no comprehensive ‘roll-up’ of information is required. Decentralizationmust be pursued as enterprises expand, otherwiseflexibility and responsiveness will erode.

One well-documented restructuring that iswidely adopted as enterprises grow is the adoptionof the multidivisional form. This involves decom-position and the devolution of decision rights

21 As Capron et al. (1998) explain, failures in the market forresources sometimes cause enterprises to buy and sell business.What they refer to as market failure appears to relate to the‘thin market’ problem discussed by the author in this paper andelsewhere (Helfat et al., 2007).

to quasi-independent profit centers. The abandon-ment of functional structures in favor of the mul-tidivisional form has been analyzed by Chan-dler (1962), Williamson (1975), and many others.The basic rationale of this reconfiguration wasto achieve greater accountability of managerialdecisions so that the recognition of opportunitiesand threats could proceed more thoroughly andexpeditiously. With functional internal structures,day-to-day problems tend to distract managementfrom long-run strategic issues. Studies showedthat decentralization along product or market lineswith independent profit centers led to performanceimprovements in many industries, at least duringthe period in which these organizational innova-tions were diffusing (Armour and Teece, 1978;Teece, 1980, 1981). More recent scholarship hassuggested that even further decentralization anddecomposition in large organizations may be ben-eficial (Bartlett and Ghoshal, 1993). There is alsosome evidence that ‘modern’ human resource man-agement techniques—involving delayering, decen-tralization of decision rights, teamwork, flexi-ble task responsibilities and performance-basedrewards—also improve performance (Jantunen,2005).

Of course, achieving decentralization can com-promise the organization’s ability to achieve inte-gration. There is little harm and much benefit fromdecentralization when the customer does not ben-efit from an integrated product offering, or whensourcing and other inputs do not benefit from inte-gration and/or aggregation. If customer and sup-ply considerations allow decomposability (becausethe required integration between units is less thanwithin units), then management’s ability to iden-tify and implement decomposable subunits shouldenhance performance. However, if firm-specificeconomies of scale and scope are available, theymust be captured—otherwise the enterprise is tan-tamount to a conglomerate. This tension can bemanaged through a collaborative nonhierarchicalmanagement style assisted by establishing councilsand other integration forums. Middle managementcan also play a critical role when such forumsare established. They can also design and imple-ment tight financial controls and performance-based reward systems. Since intangibles are keydrivers of performance, their enhancement andprotection must become a managerial priority.

The open innovation model of Chesbrough(2003) also recognizes the benefits of relying on



a distributed model of innovation where the enter-prise reaches out beyond its own boundaries toaccess and integrate technology developed by oth-ers. By way of example, Henderson and Cock-burn (1994) found that an enterprise’s ability tointegrate knowledge from external sources—their‘architectural competence’—was positively asso-ciated with research productivity, as measured bypatent counts. Likewise, Iansiti and Clark (1994)found that ‘integration capability’ in the automo-bile industry and in the computer industry wasassociated with positive enterprise performance,again demonstrating the importance of knowledgeintegration skills. In the end, it appears that infast-paced environments organizational units musthave considerable autonomy (to make decisionsrapidly) but remain connected to activities thatmust be coordinated. Achieving this delicate bal-ance is what Simon (2002) called ‘near decom-posability’ and implementing it is an importantmicrofoundation of dynamic capabilities.

Managing cospecialization

The field of strategic management and the dynamiccapabilities framework recognizes that ‘strategicfit’ needs to be continuously achieved. However,unless the concept is operationalized it has lim-ited utility. The key dimension of ‘fit’ emphasizedin the dynamic capabilities framework is that of‘cospecialization.’ The concept of cospecialization,introduced in Teece (1986) and discussed in the‘Managing complements and “platforms”’ sectionabove, operationalizes at least one dimension ofthe otherwise rather vague concept of organiza-tional adaptation and ‘fit.’ Cospecialization can beof one asset to another, or of strategy to struc-ture, or of strategy to process. It is important toboth seizing and reconfiguration. In environmentsof rapid change, there is a need for continuous orat least semi-continuous realignment.

In many ways, much of the traditional literatureon organizational adaptation and ‘fit’ (e.g., Milesand Snow, 1994) is consistent with dynamic capa-bilities. In particular, both the strategy and organi-zational behavior literature emphasize fit betweenand amongst strategy, structure, and processes.While it is not central to his framework, MichaelPorter does note that:

[S]trategic fit among many activities is fundamentalnot only to competitive advantage but also to sus-tainability of that advantage. It is harder for a rival

to match an array of interlocked activities than it ismerely to imitate a particular sales force approach,match a process technology, or replicate a set ofproduct features. (Porter, 1996: 73)

Despite Porter’s clear recognition of the conceptof ‘fit,’ neither complementarities nor cospecializa-tion are recognized in the Five Forces framework.However, complementarities and cospecializationare recognized in various ways in Teece (1986,2000, 2006, 2007), Brandenburger and Nalebuff(1996), and Santoro and McGill (2005). Economichistorians (Rosenberg, 1982; Hughes, 1983) havealso noted the phenomenon at a general level; butin most analyses of competition and competitiveadvantage, it is common to stress that various inno-vations are substitutes, rather than complementsthat may be cospecialized to each other. Indeed,Schumpeter (1934) stressed that successful inno-vations/enterprises are threatened by swarms ofimitators, all striving to produce ‘me-too’ substi-tutes.22 He completely neglected complementari-ties.

However, complementary innovation and com-plementary assets are of great significance, partic-ularly in industries in which innovation might becharacterized as cumulative, and/or where indus-try ‘platforms’ exist or are needed. Examples ofcomplementary innovation are ubiquitous. In theenterprise software industry, business applicationscan be especially valuable to users if they cansomehow be integrated into a single program, orinto a tightly integrated suite. The developmentof gyroscopic stabilizers made imaging devicessuch as video cameras and binoculars easier touse by minimizing the impact of camera shake,and enhanced the product, especially when thenew feature was able to be introduced at low cost.Likewise, better high-energy rechargeable batteriesenable laptop computers and cell phones to operatefor longer times. Situations of complementaritieswhere there is also cospecialization between tech-nologies, and between technologies and other parts

22 Schumpeter wrote (1934: 223) that innovations/new combi-nations carried out by entrepreneurs ‘are not, as one wouldexpect according to general principles of probability, evenlydistributed through time . . . but appear, if at all, discontinu-ously in groups or swarms.’ This ‘swarming’ of innovationsand innovative activity occurs ‘exclusively because the appear-ance of one or a few entrepreneurs facilitates the appearanceof others, and these the appearance of more, in even increasingnumber’ (Schumpeter, 1934: 228). Recent studies that analyzepatent races have also reinforced the view that innovations aresubstitutes, not complements.


D. J. Teece

of the value chain, are common, yet until recentlypoorly analyzed in economic analysis and in strat-egy formulation.

Cospecialized assets are a particular class ofcomplementary assets where the value of an assetis a function of its use in conjunction with otherparticular assets.23 With cospecialization, joint useis value enhancing.24 Cospecialization results in‘thin’ markets; i.e., the assets in question areidiosyncratic and cannot be readily bought andsold in a market. Capturing cospecialization ben-efits may require integrated operations (Teece,1980). Cospecialization allows differentiated prod-uct offerings or unique cost savings. The inher-ent ‘thin’ market environment surrounding spe-cific assets means that competitors are not able torapidly assemble the same assets by acquisition,and hence cannot offer the same products/servicesat competing price points.

Management’s ability to identify, develop, andutilize in combination specialized and cospecial-ized assets built or bought is an important dynamiccapability, but it is not always present in enterprisesettings. Special value can be created (and poten-tially appropriated by another party) through assetcombinations, particularly when an asset owneris not cognizant of the value of its assets toanother party that owns assets whose value willbe enhanced through combination.25 This arisesbecause the markets for cospecialized assets arenecessarily thin or nonexistent. Langlois (1992)highlights the case of the diesel-electric locomo-tive where, in the 1920s, Charles Kettering haddeveloped advanced lightweight diesel technology

23 Lippman and Rumelt’s (2003a, 2003b) recent work on devel-oping the microfoundations for resource-based theory is verycomplementary to my development of the microfoundations ofdynamic capabilities. I acknowledge their efforts in modelingcospecialized and complementary assets. In particular, they usethe concept of supermodularity to bring in the tools of cooper-ative game theory. The idea of supermodularity was introducedby Donald Topkins as a way to formalize complementarity, andis also used by economists such as Milgrom and Roberts (see, inparticular, Milgrom and Roberts, 1990) and evolutionary gametheorists to model (strategic) complementarities (for instance, inmodels of R&D spillovers).24 Complete cospecialization is a special case of economies ofscope where not only are complementary assets more valuablein joint use than in separate use, but they may, in fact, have zerovalue in separate use and high value in joint use. Cospecializationmay stem from economies of scope, but they could also stemfrom the revenue enhancement associated with producing abundled or integrated solution for the customer.25 Even if they are cognizant, they do not have the bargainingpower to take advantage of the situation.

at the GM labs. The earliest use was in sub-marines. Alfred P. Sloan, GM’s chairman, sawthe possibility of applying the technology to makediesel-electric locomotives—steam power was, atthe time, completely dominant. To accomplish this,GM needed capabilities resident in the locomotivemanufacturers and at Westinghouse Electric. Lan-glois (1992: 115) notes that the three sets of capa-bilities might have been combined by some kindof contract or joint venture, but the steam man-ufacturers—Alco, Baldwin, and Lima—failed tocooperate.26 In short, both innovation and reconfig-uration may necessitate cospecialized assets beingcombined by management in order for (systemic)innovation27 to proceed. Managers do not alwayssucceed in doing so, sometimes because they donot sense the need or the opportunity, and some-times because they do but they are unable to effec-tuate the integration. If the assets cannot be pro-cured externally, they will need to be built inter-nally.

The ability of management to identify needsand opportunities to ‘invest’ in cospecializedassets (through its own development or astutepurchase) is fundamental to dynamic capabilities.Mere ‘horse-trading’ skills (which market agentspossess) will not suffice to build sustainablecompetitive advantage, and decisions on when andhow to invest—whether and when to build orbuy cospecialized assets—will depend upon manyfactors, including transaction costs. In particular,it will depend on management’s entrepreneurialcapacities with respect to matching up andintegrating relevant cospecialized assets.

It is apparent that cospecialization involves‘lock-in’ and is a particular form of complemen-tarity that exists when technologies and otherassets need to be part of a tightly integrated sys-tem to achieve the performance that customerswant. Business success in such circumstancesrequires the coordination of R&D investment andalliance activity. The manner and timing withwhich such coordination needs to be accomplishedis important to success (Teece, 1986; Mitchell,

26 This was not because the companies feared holdup in the faceof highly specific assets. Rather, it was because they activelydenied the desirability of the diesel and fought its introductionat every step. GM was forced to create its own capabilities inlocomotive manufacturing.27 For a discussion of systemic innovation, see Teece (1988,2000).



1991). Common ownership of the parts facili-tates system-wide innovation and economic per-formance (Teece, 2000) and protects against oppor-tunism (Williamson, 1975).

To summarize, entrepreneurs and managers cancreate special value by combining cospecializedassets inside the enterprise (Teece, 2007). This mayrequire investments to create the necessary cospe-cialized technologies—as illustrated by ThomasEdison and the creation of electric power asa system. It is not uncommon in technology-based industries to find that certain technologiesare worth more to some market participants thanto others, based on the technology they alreadyhave, and their technology and product strat-egy.

Learning, knowledge management, and corporategovernance

With intangible assets being critical to enterprisesuccess, the governance and incentive structuresdesigned to enable learning and the generation ofnew knowledge become salient. There are manytypes of learning—including experiential, vicari-ous, individual, and organizational—and a largeliterature that explores each type. Also ‘sensing’requires learning about the environment and aboutnew technological capabilities. R&D was seen asone way that the enterprise could promote suchlearning. However, in the context of the dynamiccapability discussed in this section, the ability tointegrate and combine assets including knowledgeis a core skill (Kogut and Zander, 1992; Grant,1996). The combination of know-how within theenterprise, and between the enterprise and organi-zations external to it (e.g., other enterprises, uni-versities), is important.

Integrating know-how from outside as wellas within the enterprise is especially importantto success when ‘systems’ and ‘networks’ arepresent. Good incentive design and the creationof learning, knowledge-sharing, and knowledge-integrating procedures are likely to be critical tobusiness performance, and a key (micro)foundationof dynamic capabilities (Nonaka and Takeuchi,1995; Chesbrough, 2003). Of equal importanceare monitoring and managing the ‘leakage,’ mis-appropriation, and misuse of know-how, tradesecrets, and other intellectual property. Of course,tacit know-how is difficult to imitate and has acertain amount of ‘natural’ protection. However,

much know-how does leak out. Innovating busi-ness enterprises with limited experience have beenknown to inadvertently compromise or lose theirintellectual property rights. Failure to proactivelymonitor and protect know-how and intellectualproperty is common.

The outsourcing of production and the prolif-eration of joint development activities likewisecreate requirements that enterprises develop gov-ernance procedures to monitor the transfer of tech-nology and intellectual property. Technology trans-fer activities, which hitherto took place inside theenterprise, increasingly take place across enterpriseboundaries. The development of governance mech-anisms to assist the flow of technology while pro-tecting intellectual property rights from misappro-priation and misuse are foundational to dynamiccapabilities in many sectors today. Figure 3 sum-marizes the microfoundations of this third class ofdynamic capability.

There are also several other ‘governance’ issuesrelevant to dynamic capabilities. At one level thereare governance and business model issues associ-ated with an enterprise’s ability to achieve asset‘combinations’ and reconfiguration. As noted ear-lier, there is a continuous need to modify productofferings, business models, enterprise boundaries,and organizational structures. Decentralized struc-tures that facilitate near decomposability are likelyto assist in achieving reconfiguration.

One class of governance issues relate to incen-tive alignment. The microfoundations of incentiveissues are embedded in an understanding of agencyand incentive design issues, also discussed earlier.Agency theory has long emphasized that the sep-aration of ownership from control creates interestalignment problems, particularly around manage-ment compensation and the allocation of corpo-rate perquisites. The abuse of discretion and theuse of corporate assets for private purposes canoccur absent appropriate accountability/oversight.These issues become more severe as an enter-prise grows and the separation between ownershipand management widens. Recent corporate gover-nance scandals in the United States, Europe, andJapan indicate the need for continued vigilance.However, increasing the mix of independent and‘inside’ directors will not necessarily ameliorateproblems associated with strategic ‘malfeasance.’

There are likely to be benefits associated withparticipation at the board level by individuals whocan calibrate whether the top management team


D. J. Teece

Governance

Achieving Incentive Alignment;Minimizing Agency Issues;Checking Strategic Malfeasance;Blocking Rent Dissipation.

Knowledge Management

Learning;Knowledge Transfer;Know-how Integration;Achieving Know-how and IntellectualProperty Protection.

Cospecialization

Managing Strategic Fit So ThatAsset Combinations Are ValueEnhancing.

Continuous Alignment andRealignment of Specific Tangible

and Intangible Assets.

Adopting Loosely Coupled Structures;Embracing Open Innovation;Developing Integration andCoordination Skills.

Decentralization and NearDecomposability

Figure 3. Combination, reconfiguration, and asset protection skills

is sufficiently ‘dynamic.’ The replacement of theCEO and other members of the top managementteam, if they demonstrate weak sensing, seizing,and reconfiguration capabilities (strategic ‘malfea-sance’), is important to effectuate. That is not tosay that guarding against financial malfeasance isunimportant. It will always remain as an importantcorporate governance function; but its significanceis likely to pale next to strategic ‘malfeasance,’which is harder to detect and evaluate. The cur-rent wave of governance reforms in the UnitedStates—with its strong emphasis on accountingcontrols and systems integrity—may inadvertentlylead to much bigger ‘strategic’ performance fail-ures by management. Boards stacked with inex-perienced ‘independent’ board members may nothave the requisite talents to properly diagnosestrategic ‘malfeasance’ and respond accordingly.

A related literature in economics has stressedhow poorly designed incentives can produce ten-sions between the actions of employees and theactions needed to achieve profitable performance.Dysfunctional behavior, such as activity that gen-erates influence costs, has received considerableattention (Teece, 2003). Also, through use of col-lective bargaining, employees in industries insu-lated from global competition have been ableto appropriate economic surplus. Above-marketwages—which characterized, and to some extentstill characterize, certain enterprises in the auto,steel, and airline industries in the United States—are a case in point. These conditions can extend tomanagerial ranks as well. Restructuring may thenrequire the judicious use of bankruptcy laws to

rewrite uncompetitive supply contracts that are theproduct of unrealistic collective bargaining actionsin an earlier period. The ability of some enterprisesto craft work specifications, attract and retain morecommitted talent, design reward systems, developcorporate cultures, and blunt the formation ofcoalitions that extract quasi-rents through threat-ening to withhold participation, is an importantmanagerial capacity.

The design and creation of mechanisms insidethe enterprise to prevent the dissipation of rents byinterest groups (both management and employees)would also appear to be very relevant to dynamiccapabilities, but has not been high on the agenda ofstrategy researchers. One exception is Gottschalgand Zollo (2007), who point out that the capac-ity to continuously achieve incentive alignmentis an important performance-enhancing (and rent-protecting) dynamic capability.

Many of the issues discussed here have, in thepast, fallen under the rubric of human resourcemanagement; a closer connection of these issuesto strategic management issues would appear tobe warranted. The reason is that strategic man-agement is focused not only on how to gen-erate rent streams, but also on how to preventthem from being dissipated or captured by variousentities or groups inside and outside the enter-prise. For instance, the concepts of the ‘appropri-ability regime’ and ‘isolating mechanisms’ weredeveloped by strategic management scholars tohelp explain how rents from innovation and othersources of superior performance can be protectedand guarded from dissipation by competitors and



others. However, the earlier focus on marketsor ‘external’ competition did not address internalappropriation by interest groups.

DYNAMIC CAPABILITIES,‘ORCHESTRATION’ SKILLS, ANDCOMPETITIVE ADVANTAGE

The general framework advanced here seesdynamic capabilities as the foundation of enter-prise-level competitive advantage in regimes ofrapid (technological) change. The framework indi-cates that the extent to which an enterprise devel-ops and employs superior (nonimitable) dynamiccapabilities will determine the nature and amountof intangible assets it will create and/or assembleand the level of economic profits it can earn (seeFigure 4). Furthermore, the framework emphasizesthat the past will impact current and future perfor-mance. However, there is much that managementcan do to simultaneously design processes andstructures to support innovation while unshacklingthe enterprise from dysfunctional processes andstructures designed for an earlier period.

In Teece and Pisano (1994) and Teece et al.(1997), we proposed three organizational and man-agerial processes—coordination/integrating, learn-ing, and reconfiguring—as core elements ofdynamic capabilities. These processes are a sub-set of the processes that support sensing, seiz-ing, and managing threats. Together they mightbe thought of as asset ‘orchestration’ processes.A key strategic function of management is tofind new value-enhancing combinations inside theenterprise, and between and amongst enterprises,and with supporting institutions external to theenterprise. Because many of the most valuableassets inside the firm are knowledge related andhence nontradable, the coordination and integra-tion of such assets create value that cannot be repli-cated in a market. This establishes a distinctive rolefor managers in economic theory and in the eco-nomic system. Managers seek new combinationsby aligning cospecialized assets (Teece, 2007). Theneed to sense and seize opportunities, as well asreconfigure when change occurs, requires the allo-cation, reallocation, combination, and recombina-tion of resources and assets. These are the keystrategic function of executives. Indeed, skills usedto identify and exploit complementarities and man-age cospecialization are scarce. Figuring out how

to increase value from the use of the assets theenterprise owns involves knowing the fine-grainedstructure of the firm’s asset base, and filling inthe gaps necessary to provide superior customersolutions. Gap filling may involve building newassets, or acquisitions and strategic partnerships(Ettlie and Pavlou, 2006).

The dynamic capabilities framework recognizesthat the business enterprise is shaped but notnecessarily trapped by its past. Management canmake big differences through investment choiceand other decisions. Enterprises can even shapetheir ecosystem. In this sense, the frameworkis quite Chandlerian (Chandler, 1990a, 1990b).Managers really do have the potential to settechnological and market trajectories, particularlyearly on in the development of a market (David,1992). Indeed, the enterprise and its environ-ment frequently coevolve. However, because of theassumed context—regimes of rapid technologicalchange exposed to the full force of internationalcompetition—there is little room for big mistakes.

Hence, the dynamic capabilities framework ispartially but not entirely in the spirit of evolution-ary theorizing. The dynamic capabilities frame-work endeavors to capture the key variables andrelationships that need to be ‘manipulated’ to cre-ate, protect, and leverage intangible assets so as toachieve superior enterprise performance and avoidthe zero-profit trap. However, building and assem-bling tangible and intangible assets and effectuat-ing change is seen as difficult. Long-run successis likely to require achieving necessary internalcreative destruction, possibly involving spin-outsand spin-offs to help sustain superior performance.Decision biases must also be neutralized. In short,enterprises may be more like biological organ-isms than some economists, managers, and strategyscholars are willing to admit; but they are alsomore malleable than some organizational ecolo-gists are willing to recognize.

The enterprise will need sensing, seizing, andtransformational/reconfiguring capabilities to besimultaneously developed and applied for it tobuild and maintain competitive advantage. Simul-taneity may not be necessary at the product level—indeed, Helfat and Peteraf (2003) distinguishbetween capability development and subsequenthoning, grafting, and branding. Endeavoring tosimultaneously achieve sensing, seizing, and recon-figuring at the individual product level could leadto chaos and lack of effectiveness, as routines and


D. J. Teece

SE

NS

ING

SE

IZIN

GM

AN

AG

ING

TH

RE

AT

S/

TR

AN

SF

OR

MIN

G

DY

NA

MIC

CA

PA

BIL

ITIE

S

SE

LEC

TE

DM

ICR

O-

FO

UN

DA

TIO

NS

Sel

ectin

gD

ecis

ion-

Mak

ing

Pro

toco

ls

Del

inea

ting

the

Cus

tom

erS

olut

ion

and

the

Bus

ines

s M

odel

Bui

ldin

g Lo

yalty

and

Com

mitm

ent

Sel

ectin

gE

nter

pris

eB

ound

arie

s to

Man

age

Com

plem

ents

and

“Con

trol

”P

latfo

rms.

Pro

cess

es to

Tap

Sup

plie

r and

Com

plem

ento

rIn

nova

tion.

Pro

cess

es to

Dire

ct In

tern

alR

&D

and

Sel

ect

New

Tec

hnol

ogie

s.

Pro

cess

es to

Iden

tify

Tar

get

Mar

ket

Seg

men

ts,

Cha

ngin

gC

usto

mer

Nee

dsan

d C

usto

mer

Inno

vatio

n.

Pro

cess

es to

Tap

Dev

elop

men

ts in

Exo

geno

usS

cien

ce a

ndT

echn

olog

y.

Gov

erna

nce

Dec

entr

aliz

atio

nan

d N

ear

Dec

ompo

sabi

lity

Kno

wle

dge

Man

agem

ent.

Cos

peci

aliz

atio

n.

Ent

erpr

ise

Str

uctu

res,

Pro

cedu

res,

Des

igns

and

Ince

ntiv

esfo

r S

eizi

ngO

ppor

tuni

ties

Con

tinuo

usA

lignm

ent a

ndR

ealig

nmen

t of

Spe

cific

Tan

gibl

e an

dIn

tang

ible

Ass

ets

Ana

lytic

al S

yste

ms

(and

Indi

vidu

alC

apac

ities

) to

Lear

nan

d to

Sen

se, F

ilter

,S

hape

, and

Cal

ibra

teO

ppor

tuni

ties.

Figu

re4.

Foun

datio

nsof

dyna

mic

capa

bilit

ies

and

busi

ness

perf

orm

ance



rules in the organization would likely be in a con-tinuous state of flux.

The first two capabilities recognized as funda-mental—sensing and seizing—are related to butdifferent from March’s (1991, 1996, 2006) con-cepts of exploration and exploitation. March seemsclear that both are necessary for adaptation, buthe has recognized the tensions, if not incompat-ibilities, between the two. His argument in partis that incompatibilities flow from the fact thatexploration and exploitation compete for resourcesand that the mindsets and organizational routinesneeded for one are different from the other, mak-ing simultaneous pursuit difficult, if not impossible(March, 1996: 280). While there is merit to eachassumption, both need to be put into perspective.With respect to competition for resources, sensingdoes not necessarily involve large commitmentsof resources, at least not relative to seizing. Cer-tain aspects, such as monitoring the environment,can be a low-cost activity. Early-stage exploratoryresearch is also relatively inexpensive. Mansfieldet al.’s (1971: Table 6.2) studies of new prod-uct development showed that the cost of early-stage research activities was a small percentageof the total new product development costs. Forinstance, the costs of pharmaceutical developmenttypically far exceed those of pharmaceutical dis-covery. Also, with respect to the different mindsetsand routines, while there are undoubtedly tensions,these can be relieved by having different organi-zation units (or different parts of an organizationalunit) specializing to some degree on sensing ascompared to seizing. As Gupta, Smith, and Shal-ley (2006: 697) note: ‘exploration or exploitationin one domain may coexist with high levels ofexploration or exploitation in the other domain.’Of course, the outsourcing of manufacturing andother aspects of seizing reconciles the issues evenmore starkly, as the routines needed for proficientmanufacturing then lie external to the firm.

The need for both exploration and exploitationis well accepted for adaptative systems, and isembedded in the literature on ambidexterity (e.g.,O’Reilly and Tushman, 2007). This literature rec-ognizes that both exploration and exploitation canbe assisted by differentiated but partially or weaklyintegrated subunits (divisions, departments). Sens-ing activities need to be decentralized with theinformation rolling up to top management. Tightplanning will be a part of seizing, but less so ofsensing.

To summarize, an enterprise’s ability to man-age competitor threats and to reconfigure itselfis dependent on its investment activity, whichis in turn dependent on its ability to sense anopportunity. This aspect of dynamic capabilitiesindicates that the likelihood of achieving finan-cial success depends on events and responses tothem. Formally, let the probability of a high eco-nomic profits ranking for an enterprise, condi-tional on some extraordinary event E (e.g., anexogenous technological change that opens up thepossibility of a new business opportunity) occur-ring,28 be Pr(�|E). Then: Pr(�|E = Pr(sense|E) ×Pr(seize|E, sense) × Pr(manage threats/transform|E, sense, seize) × P r((�|E), sense, seize,manage threats/transform).

As indicated throughout this paper and through-out earlier treatments by this author, it is alsonecessary to assess the issue of the ‘sustainabil-ity’ or nonimitability of both assets and capa-bilities. This in turn depends upon a number offactors summarized adequately by the twin con-cepts of ‘isolating mechanism’ and ‘appropriabil-ity regimes.’29 When the appropriability regime is‘tight’ and the business enterprise’s own isolatingmechanisms are strong, differential performancecan be sustained, at least for a time. Dynamiccapabililties of course require the creation, integra-tion, and commercialization of a continuous streamof innovation consistent with customer needs andtechnological opportunities.

Note that in the dynamic capabilities framework,enterprises must employ sensing, seizing, andreconfiguring mechanism to direct their financialresources consistent with marketplace needs andimperatives. However, as a matter of pure theory,enterprises need not continuously reinvent them-selves. The need to reinvent depends on events,anticipated or otherwise. If the ecosystem in whichthe enterprise is embedded remains stable, the needto change can be modulated accordingly.30 Indeed,

28 Alternatively, one could assess the unconditional probabilityPr(�) of earning such profits. Pr(�) = Pr(�|E) + Pr(�| ∼E) with Pr(�| ∼ E) defined analogously to the definition ofPr(�|E) in the text. In competitive markets without dynamiccapabilities, Pr(�| ∼ E) is likely to be zero.29 Intellectual property protection, the tacit nature of know-how,and the inherent difficulty of the technology, all affect the ease ofimitation. Another factor developed in this article is the uniquecoalignment of specific assets. Achieving such combinationsmay be difficult for imitators to effectuate.30 This assumes that the ecosystem remains attractive. If it doesnot, the enterprise will have to consider migrating to a different


D. J. Teece

if an enterprise controls standards, or can some-how help stabilize its own environment, then itmay not need to engage in the continuous andcostly exploration of radical alternatives (March,1991). Selecting suitable business models, makingthe right strategic investment decisions, and pursu-ing incremental innovation can keep an enterprisehighly competitive for a decade or so (e.g., Boe-ing’s decision to build the 747, which 30 yearslater is much improved and still competitive insome configurations on some routes) if the envi-ronment is stable. Excessive internal change forthe sake of it can lead to internal chaos and per-formance failure.

RESOURCES/COMPETENCESDISTINGUISHED FROM DYNAMICCAPABILITIES

The dynamic capabilities framework advances aneo-Schumpeterian theory of the firm and orga-nizational decision making that is recognizable tothose familiar with the behavioral theory of thefirm, with evolutionary theorizing in economics,and with a Schumpeterian characterization of theinnovation process. It also builds on what has cometo be known as the resource-based approach. Whilethe resource-based approach is inherently static, itis nevertheless relevant to dynamic capabilities. Asnoted by Teece et al.:

the resource-based perspective also invites consid-eration of strategies for developing new capabil-ities. Indeed, if control over scarce resources isthe source of economic profits, then it followsthat such issues as skill acquisition and learningbecome fundamental strategic issues . . . (Teeceet al., 1990a: 9)

Zott similarly recognizes that

dynamic capabilities are more than a simple addi-tion to resource based view since they manipulatethe resources and capabilities that directly engenderrents. (Zott, 2003: 120)

Collis (1994) and Winter (2003) also note thatone element of dynamic capabilities is that they

ecosystem, or reshaping the ecosystem itself. Both are verychallenging tasks.

govern the rate of change of ordinary capabili-ties.31 However, the notion advanced here is that,at least analytically, dynamic capabilities can bedisaggregated into sensing, seizing, and transfor-mational activities. Enterprises with good dynamiccapabilities will have entrepreneurial managementthat is strategic in nature and achieves the value-enhancing orchestration of assets inside, between,and amongst enterprises and other institutionswithin the business ecosystem. Dynamic capabilityis a meta-competence that transcends operationalcompetence. It enables firms not just to invent butalso to innovate profitably (Teece, 1986, 2006).

The dynamic capabilities framework is integra-tive. Dosi, Nelson, and Winter (2000: 4) noted atone point the ‘terminological flotilla’ in the lit-erature on organizational competences. However,perhaps there is now an emerging consensus thatresources/competences map well into what his-torically we have thought of as the enterprise’soperational capabilities, which help sustain techni-cal fitness. Dynamic capabilities, by contrast, relateto high-level activities that link to management’sability to sense and then seize opportunities, nav-igate threats, and combine and reconfigure spe-cialized and cospecialized assets to meet changingcustomer needs, and to sustain and amplify evolu-tionary fitness, thereby building long-run value forinvestors.

If an enterprise possesses resources/competencesbut lacks dynamic capabilities, it has a chanceto make a competitive return (and possibly evena supra-competitive return) for a short period;but it cannot sustain supra-competitive returns forthe long term except due to chance. It may earnRicardian (quasi-)rents when demand increases forits output, but such quasi-rents will be competedaway. It does not earn those Schumpeterian rentsassociated with ‘new combinations’ and subse-quent recombination, or Kirznerian rents associ-ated with bringing markets back into equilibrium.It might earn short-term Porterian rents associatedwith ‘building defenses against competitive forces’(Porter, 1991: 22), but this is far too reactive forlong-term success. Dynamically competitive enter-prises don’t just build defenses to competition;they help shape competition and marketplace out-comes through entrepreneurship, innovation, and

31 As discussed here, dynamic capabilities certainly include thiselement, as well as several others.



semi-continuous asset orchestration and businessreconfiguration.

The archetypical enterprise with competences/resources but lacking dynamic capabilities willin equilibrium ‘earn a living by producing andselling the same product, on the same scale andto the same customer population’ (Winter, 2003:992). Such an enterprise might even be goodat invention, but it will likely fail to capitalizeon its technological accomplishments. The oper-ational/technical competences possessed might in-clude basic ones such as order entry (to commu-nicate what needs to be made/supplied), billings(to collect from customers), purchasing (to decidewhat inputs to buy and to then pay suppliers),financial controls (to restrict behavior and preventtheft), inventory controls (to minimize inventorycosts) financial reporting (to access capital), mar-keting (to identify customers), and sales (to obtainorders). Management of these functions is com-monly considered operations management.

Operations management is arguably at the foun-dation of basic management functions; but whileknowledge of modern production systems tookgenerations to develop, it is now widely diffused.The division of labor, uniform standards, the mov-ing assembly line, measurement techniques forinspection, and control all of course had to beinvented and they now constitute what we nowthink of as the (American) system of production.

Competitive advantage can in theory flow fromsuperior operations, or what was referred to ear-lier as ‘technical fitness.’ Indeed, the IndustrialRevolution saw significant differentials open upbetween craft systems and modern production sys-tems, and these innovations led to an almost com-plete reordering of the industrial landscape. AsCharles Babbage noted almost 200 years ago:

[W]e shall notice, in the art of making even themost insignificant of [articles], processes calculatedto excite our admiration by their simplicity, or torivet our attention by their unlooked-for results.(Babbage, 1835: 3)

However, the postwar period has led to greatprogress in the understanding of how produc-tion systems work. Many useful techniques havebeen developed and improved. With developmentsin the field of management science and opera-tions research, precise answers to narrow problemsexist. Much is known about inventory manage-ment, scheduling, planning, quality control, and

about managing isolated subsystems. The pursuitof ‘benchmarking’ and the adoption of ‘best prac-tices’ has helped with the diffusion of discreteskills, protocols, and procedures. However, accord-ing to one of the field’s pioneers, ‘we have notlearned very much about the relationships betweenthese subsystems’ (Buffa, 1982: 2). This is oneplace where dynamic capabilities come into play.

One implication is that special know-how—know-how that is difficult to obtain and apply—isneeded to sense opportunities, execute plans, andconfigure and reconfigure assets and systems asnecessary. Skill in putting things together to cap-ture cospecialization benefits is important. Evenwith respect to operations management, it seemsthe pay-off today is in understanding how sub-systems are related and interact together. Put dif-ferently, the understanding of the basic businessfunctions that constitute business administrationand operations management is widely diffused andhence well known, at least in advanced economies.The wide diffusion of knowledge with respect tosuch functions means that much can be outsourcedor implemented inside any enterprise with rela-tive facility. However, by running hard at this, anenterprise may manage only to stand still—whatsome refer to as the ‘Red Queen’ effect. Absenta broader overarching set of dynamic capabilities,a firm that is merely competent in operations willfail. However, understanding how to enhance per-formance of the enterprise through sensing futureneeds, making quality, timely, and unbiased invest-ment decisions inside a well-designed businessmodel, executing well on those decisions, effec-tuating productive combinations, promoting learn-ing, reengineering systems that no longer workwell, and implementing good governance remainsenigmatic. The requisite managerial services thatundergird dynamic capabilities cannot be out-sourced. Understanding and implementing the pro-cesses and structures that undergird dynamic capa-bilities is enterprise specific, and requires intimateknowledge of both the enterprise and the ecosys-tem in which the enterprise cooperates and com-petes.

In this regard, a useful distinction can be madebetween entrepreneurs, managers, and administra-tors. Administrators are responsible for the day-to-day operations and the routine; they help ensurethat the enterprise is technically fit, in the sensedefined earlier. They are not expected to engage inentrepreneurial activities; e.g., they are not relied


D. J. Teece

on to sense new business opportunities. Nor arethey typically expected to discover the need forand to design new enterprise-wide operating rou-tines, as this constitutes evolutionary fitness. Thedistinctions made earlier are implicitly recognizedby Porter (1996) when he claims that operationaleffectiveness is not strategy. He recognizes thatboth operational effectiveness and strategy areessential to superior performance, but notes:

The quest for productivity, quality, and speed hasspawned a remarkable number of managementtools and techniques, total quality managementbenchmarking, time-based competition, outsourc-ing, partnering, reengineering and change manage-ment. Although the resulting operational improve-ments have been dramatic, many companies havebeen frustrated by their inability to translate gainsinto sustainable profitability. And bit-by-bit, almostimperceptibly, management tools have taken theplace of strategy. As managers push to improveon all fronts, they move farther away from viablecompetitive positions (Porter, 1996: 61).

Yet it is perhaps an overstatement to say that ‘oper-ations management’ tools and procedures cannotbe the basis of competitive advantage, or workagainst it. If there is a significant, tacit, non-inimitable component of an enterprise’s superioroperational competence, it has the potential fora time to support superior performance (it will,in fact, generate Ricardian rents).32 Nevertheless,superior operational efficiency, while valuable, isnot a dynamic capability.

CONCLUSION

For open economies exposed to rapid technolog-ical change, the dynamic capabilities frameworkhighlights organizational and (strategic) manage-rial competences that can enable an enterprise toachieve competitive advantage, and then semi-continuously morph so as to maintain it. Theframework integrates and synthesizes concepts andresearch findings from the field of strategic man-agement, from business history, industrial eco-nomics, law and economics, the organizational sci-ences, innovation studies, and elsewhere.

Implicit in the dynamic capabilities frameworkis a recognition that relatively open regimes of

32 Wal-Mart and Dell Inc. have both used differentiated businessmodels to anchor their competitive advantages.)

free trade and investment, global dispersion in thesources of new knowledge, and the multi-inventionor systemic character of such innovation have‘upped the ante’ for modern management. Improv-ing quality, controlling costs, lowering inventories,and adopting best practices (’technical fitness’)will no longer suffice for long-run competitivesuccess. Nor do traditional scale economies in pro-duction always have the differentiating power theymay once have had. More than scale and scopeadvantage are needed. Success requires the cre-ation of new products and processes and the imple-mentation of new organizational forms and busi-ness models, driven by an intensely entrepreneurialgenre of management constantly honing the evo-lutionary and entrepreneurial fitness of the enter-prise. Entrepreneurial managers can sense andeven help shape the future, unshackle the enter-prise from the past, and stay ahead by augment-ing knowledge assets, protecting them with intel-lectual property rights, establishing new value-enhancing asset combinations, and transformingorganizational and, if necessary, regulatory andinstitutional structures. Dynamic capabilities residein large measure with the enterprise’s top manage-ment team, but are impacted by the organizationalprocesses, systems, and structures that the enter-prise has created to manage its business in thepast.

Maintaining dynamic capabilities thus requiresentrepreneurial management. The entrepreneurialmanagement in question is different but relatedto other managerial activity. Entrepreneurship isabout sensing and understanding opportunities,getting things started, and finding new and betterways of putting things together. It is about cre-atively coordinating the assembly of disparate andusually cospecialized elements, getting ‘approvals’for nonroutine activities, and sensing businessopportunities. Entrepreneurial management has lit-tle to do with analyzing and optimizing. It is moreabout sensing and seizing—figuring out the nextbig opportunity and how to address it.

We have come to associate the entrepreneurwith the individual who starts a new businessproviding a new or improved product or ser-vice. Such action is clearly entrepreneurial, butthe entrepreneurial management function embed-ded in dynamic capabilities is not confined tostartup activities and to individual actors. It isa new hybrid: entrepreneurial managerial capital-ism. It involves recognizing problems and trends,



directing (and redirecting) resources, and reshap-ing organizational structures and systems so thatthey create and address technological opportu-nities while staying in alignment with customerneeds. The implicit thesis advanced here is thatin both large and small enterprises entrepreneurialmanagerial capitalism must reign supreme forenterprises to sustain financial success. Nor isentrepreneurial management merely ‘intrapreneur-ship,’ as there is a large role for the entrepreneurialmanager in external activities, including shapingthe ecosystem.

As discussed, there are obvious tensions andinterrelationships between and amongst the threeclasses of capabilities identified. The managerialskills needed to sense are quite different from thoseneeded to seize and those needed to reconfigure.All functions have a significant ‘entrepreneurial’and ‘right brain’ component. Successful enter-prises must build and utilize all three classes ofcapabilities and employ them, often simultane-ously. Since all three classes are unlikely to befound in individual managers, they must be some-where represented in top management, and theprincipal executive officer must succeed in gettingtop management to operate as a team. Of course,if the principal executive officer has depth in allthree classes of capabilities, the organization has abetter chance of success.

The dynamic capabilities framework goesbeyond traditional approaches to understandingcompetitive advantage in that it not only empha-sizes the traits and processes needed to achievegood positioning in a favorable ecosystem, but italso endeavors to explicate new strategic consider-ations and the decision-making disciplines neededto ensure that opportunities, once sensed, can beseized; and how the business can be reconfiguredwhen the market and/or the technology inevitablyis transformed once again. In this sense, dynamiccapabilities aspire to be a relatively parsimoniousframework for explaining an extremely seminaland complicated issue: how a business enterpriseand its management can first spot the opportu-nity to earn economic profits, make the decisionsand institute the disciplines to execute on thatopportunity, and then stay agile so as to contin-uously refresh the foundations of its early success,thereby generating economic surpluses over time.If the framework has succeeded in some smallmeasure, then we have the beginnings of a gen-eral theory of strategic management in an open

economy with innovation, outsourcing, and off-shoring.

ACKNOWLEDGEMENTS

An earlier version of this paper formed the basisof a lecture delivered in Finland on September3, 2003, on the occasion of the award of theViipuri Prize in Strategic Management. I wish tothank Mie Augier, John Blair, Hank Chesbrough,Antonia Cusumano Binette, Connie Helfat, DanLovallo, Cyndi Morrow, Richard Nelson, IkijiroNonaka, Christos Pitelis, Dan Schendel, ValeskaScheren-Guivel, Edward F. Sherry, and SydneyWinter for many helpful comments and sugges-tions on early drafts. I also thank Richard Rumeltfor insights provided over the years, which haveundoubtedly shaped my thinking on strategy andon dynamic capabilities. Two anonymous refer-ees provided trenchant comments on several dif-ferent drafts. These were immensely helpful, andI am much indebted to them. Patricia Murphy,Patricia Lonergan, Frances Darnley, and ElizabethOlson provided tenacious assistance in preparingthe manuscript for publication.

REFERENCES

Abernathy WJ, Utterback JM. 1978. Patterns of industrialinnovation. Technology Review 80(7): 40–47.

Alchian A, Demsetz H. 1972. Production, informationcosts, and economic organization. American EconomicReview 62: 777–795.

Amit R, Schoemaker PJH. 1993. Strategic assets andorganization rent. Strategic Management Journal14(1): 33–46.

Armour H, Teece DJ. 1978. Organizational structureand economic performance: a test of the multidivi-sional hypothesis. Bell Journal of Economics 9(2):106–122.

Audretsch DB. 1995. Innovation, growth, and survival.International Journal of Industrial Organization13(4): 441–457.

Babbage C. 1835. On the Economy of Machinery andManufactures (4th edn). Charles Knight: London.

Bartlett CA, Ghoshal S. 1993. Beyond the M-form:toward a managerial theory of the enterprise. StrategicManagement Journal , Winter Special Issue 14:23–46.

Baumol W. 2006. Entrepreneurship and invention: towardrestoration into microeconomic value theory. Workingpaper, Ringberg Castle Presentation, Germany.

Brandenburger AM, Nalebuff BJ. 1996. Co-opetition .Harvard Business School Press: Boston, MA.


D. J. Teece

Branzei O, Vertinsky I. 2006. Pathways to productinnovation capabilities in SMEs. Journal of BusinessVenturing 21(1): 75–105.

Buffa E. 1982. Research in operations management.Journal of Operations Management 1(1): 1–7.

Capron L, Dussauge P, Mitchell W. 1998. Resource rede-ployment following horizontal mergers and acqui-sitions in Europe and North America, 1988–1992.Strategic Management Journal 19(7): 631–661.

Casson M. 1997. Information and Organization: A NewPerspective on the Theory of the Enterprise. OxfordUniversity Press: New York.

Chandler A. 1962. Strategy and Structure: Chapters in theHistory of Industrial Enterprise. Harvard UniversityPress: Cambridge, MA.

Chandler A. 1990a. Scale and Scope: The Dynamicsof Industrial Capitalism . Harvard University Press:Cambridge, MA.

Chandler A. 1990b. The enduring logic of industrialsuccess. Harvard Business Review 68(2): 130–140.

Chesbrough H. 2003. Open Innovation: The New Impera-tive for Creating and Profiting from Technology . Har-vard Business School Press: Boston, MA.

Chesbrough H, Rosenbloom RS. 2002. The role of thebusiness model in capturing value from innovation:evidence from Xerox Corporation’s technology.Industrial and Corporate Change 11(3): 529–555.

Chesbrough H, Teece DJ. 1996. Organizing for inno-vation: when is virtual virtuous? Harvard BusinessReview 74(1): 65–73.

Collis DJ. 1994. Research note: how valuable areorganizational capabilities? Strategic ManagementJournal , Winter Special Issue 15: 143–152.

Cyert RM, March JG. 1963. A Behavioral Theory of theEnterprise. Prentice-Hall: Englewood Cliffs, NJ.

David P. 1992. Heroes, herds and hysteresis intechnological history: Thomas Edison and the battleof the system reconsidered. Industrial and CorporateChange 1(1): 129–180.

Davidow W, Malone M. 1992. The Virtual Corporation .Harper Business: New York.

Dosi G, Nelson RR, Winter SG. 2000. Introduction.In The Nature and Dynamics of OrganizationalCapabilities , Dosi G, Nelson, RR, Winter SG (eds).Oxford University Press: New York; 1–24.

Dunne T, Roberts MJ, Samuelson L. 1988. Patterns ofenterprise entry and exit in U.S. manufacturingindustries. Rand Journal of Economics 19(4):495–515.

Eisenhardt K, Martin J. 2000. Dynamic capabilities:what are they? Strategic Management Journal ,October–November Special Issue 21: 1105–1121.

Ettlie JE, Pavlou PA. 2006. Technology-based newproduct development partnerships. Decision Sciences37(2): 117–147.

Evans DS, Hagiu A, Schmalensee R. 2006. InvisibleEngines: How Software Platforms Drive Innovationand Transform Industries . MIT Press: Cambridge,MA.

Freeman C. 1974. The Economics of Industrial Innova-tion . Penguin: Harmondsworth, U.K.

Ghemawat P. 1991. Commitment: The Dynamics ofStrategy . Free Press: New York.

Gottschalg O, Zollo M. 2007. Interest alignment andcompetitive advantage. Academy of ManagementReview 32(2): 418–437.

Grant RM. 1996. Prospering in dynamically competitiveenvironments. Organizational Science 7(4): 375–387.

Gupta AK, Smith KG, Shalley CE. 2006. The interplaybetween exploration and exploitation. Academy ofManagement Journal 49(4): 693–706.

Handy C. 1990. The Age of Unreason . Harvard BusinessSchool Press: Boston, MA.

Helfat C, Peteraf M. 2003. The dynamic resource-basedview: capability lifecycles. Strategic ManagementJournal, October Special Issue 24: 997–1010.

Helfat C, Finkelstein S, Mitchell W, Peteraf MA, SinghH, Teece DJ, Winter SG. 2007. Dynamic Capabilities:Understanding Strategic Change in Organizations .Blackwell: Oxford, U.K.

Henderson RM. 1994. Managing innovation in theinformation age. Harvard Business Review 72(1):100–106.

Henderson RM, Clark K. 1990. Architectural innovation:the reconfiguration of existing product technologiesand the failure of established firms. AdministrativeScience Quarterly 35: 9–30.

Henderson RM, Cockburn I. 1994. Measuring com-petence? Exploring firm effects in pharmaceuticalresearch. Strategic Management Journal , Winter Spe-cial Issue 15: 63–84.

Hughes TP. 1983. Networks of Power Electrificationin Western Society 1880–1930 . Johns HopkinsUniversity Press: Baltimore, MD.

Iansiti M, Clark KB. 1994. Integration and dynamiccapability: evidence from product development inautomobiles and mainframe computers. Industrial andCorporate Change 3(3): 557–605.

Jacobides MG, Winter S. 2005. The coevolution of capa-bilities and transaction costs: explaining the institu-tional structure of production. Strategic ManagementJournal 26(5): 395–413.

Jantunen A. 2005. New HRM practices and knowledgeutilization. In Proceedings of the 5th InternationalWorkshop on Human Resource Management, Seville,Spain.

Kahneman D, Lovallo D. 1993. Timid choices and boldforecasts: a cognitive perspective on risk taking.Management Science 39(1): 17–31.

Kahneman D, Tversky A. 1979. Prospect theory: ananalysis of decisions under risk. Econometrica 47(2):263–291.

Kirzner I. 1973. Competition and Entrepreneurship.University of Chicago Press: Chicago, IL.

Klepper S, Graddy E. 1990. The evolution of newindustries and the determination of market structure.Rand Journal of Economics 21(1): 27–44.

Klepper S, Miller J. 1995. Entry, exit, and shakeoutsin the United States in new manufactured products.Internal Journal of Industrial Organization 13(4):567–591.

Knight FH. 1921. Risk, Uncertainty, and Profit . HoughtonMifflin: New York.



Kogut B, Zander U. 1992. Knowledge of the enterprise,combinative capabilities and the replication oftechnology. Organizational Science 3(3): 383–397.

Langlois R. 1992. Transactions-cost economics in realtime. Industrial and Corporate Change 1(1): 99–127.

Lazonick W. 2005. The innovative firm. In The OxfordHandbook of Innovation , Fagerberg J, Mowery D,Nelson RR (eds). Oxford University Press: New York;29–55.

Leonard-Barton D. 1995. Wellsprings of Knowledge-Building and Sustaining the Sources of Innovation .Harvard Business School Press: Boston, MA.

Lippman SA, Rumelt RP. 1982. Uncertain imitability: ananalysis of interfirm differences under competition.Bell Journal of Economics 13(2): 418–438.

Lippman SA, Rumelt RP. 2003a. A bargaining perspec-tive on resource advantage. Strategic ManagementJournal 24(11): 1069–1086.

Lippman SA, Rumelt RP. 2003b. The payments perspec-tive: micro-foundations of resource analysis. Strate-gic Management Journal , October Special Issue 24:903–927.

Macher JT, Mowery DC. 2004. Vertical specializationand industry structure in high technology industries.Advances in Strategic Management 21: 317–356.

Malerba F, Orsenigo L. 1996. The dynamics andevolution of industries. Industrial and CorporateChange 5(1): 51–87.

Mansfield E, Rapoport J, Schnee J, Wagner S, Ham-burger M. 1971. Research and Innovation in the Mod-ern Corporation. WW Norton: New York.

March JG. 1991. Exploration and exploitation inorganizational learning. Organizational Science 2(1):71–87.

March JG. 1996. Continuity and change in theoriesof organizational action. Administrative ScienceQuarterly 41: 278–287.

March JG. 2006. Rationality, foolishness, and adaptiveintelligence. Strategic Management Journal 27(3):201–214.

March JG, Simon HA. 1958. Organizations . Wiley: NewYork.

Miles R, Snow C. 1994. Fit, Failure, and the Hall ofFame: How Companies Succeed or Fail . Free Press:New York.

Milgrom P, Roberts J. 1990. The economics of modernmanufacturing: technology, strategy, and organization.American Economic Review 80(3): 511–528.

Mitchell W. 1991. Dual clocks: entry order influenceson industry incumbent and newcomer market shareand survival when specialized assets retain their value.Strategic Management Journal 12(2): 85–100.

Monteverde K, Teece DJ. 1982. Supplier switching costsand vertical integration in the U.S. automobileindustry. Bell Journal of Economics 13(1): 206–213.

Nelson RR. 2005. Technology, Institutions, and EconomicGrowth . Harvard University Press: Cambridge, MA.

Nelson RR, Winter SG. 1982. An Evolutionary Theoryof Economic Change. Harvard University Press:Cambridge, MA.

Nelson RR, Winter SG. 2002. Evolutionary theorizing ineconomics. Journal of Economic Perspectives 16(2):23–46.

Nonaka I, Takeuchi H. 1995. The Knowledge CreatingCompany . Oxford University Press: New York.

Nonaka I, Toyama R. 2007. Strategic management asdistributed practical wisdom (phronesis). Industrialand Corporate Change 16(3): 371–394.

O’Reilly CA, Tushman M. 2007. Ambidexterityas a dynamic capability: resolving the inno-vator’s dilemma. (March 2007). http://ssrn.com/abstract=978493 [24 June 2007].

Packard D. 1995. The HP Way: How Bill Hewlett and IBuilt Our Company . HarperCollins: New York.

Phillips A. 1971. Technology and Market Structure: AStudy of the Aircraft Industry . Heath Lexington Books:Lexington, MA.

Pisano G, Shan W, Teece DJ. 1988. Joint ventures andcollaboration in the biotechnology industry. In Inter-national Collaborative Ventures in U.S. Manufactur-ing , Mowery D (ed.). Ballinger: Cambridge, MA;183–222.

Porter M. 1980. Competitive Strategy: Techniques forAnalyzing Industries and Competitors . Free Press:New York.

Porter M. 1991. How competitive forces shape strategy.In Strategy: Seeking and Securing CompetitiveAdvantage, Montgomery C, Porter M (eds). HarvardBusiness School Press: Boston, MA; 11–26.

Porter M. 1996. What is strategy? Harvard BusinessReview 74(6): 61–80.

Rosenberg N. 1982. Inside the Black Box. CambridgeUniversity Press: New York.

Rosenkopf L, Nerkar A. 2001. Beyond local search:boundary-spanning, exploration and impact in theoptical disc industry. Strategic Management Journal22(4): 287–306.

Rumelt R. 1984. Towards a strategic theory of theenterprise. In Competitive Strategic Management ,Lamb RB (ed). Prentice-Hall: Englewood Cliffs, NJ;556–570.

Rumelt R. 1995. Inertia and transformation. In ResourceBased and Evolutionary Theories of the Enterprise,Montgomery C (ed.). Kluwer Academic: Boston, MA;101–132.

Santoro D, McGill JP. 2005. The effect of uncer-tainty and asset cospecialization on governance inbiotechnology alliances. Strategic Management Jour-nal 26(13): 1261–1269.

Schumpeter J. 1934. The Theory of Economic Develop-ment. Harvard University Press: Cambridge, MA.

Shane S. 2003. A General Theory of Entrepreneurship.Edward Elgar: Northampton, MA.

Simon HA. 1993. Altruism and economics. AmericanEconomic Review 83(2): 156–161.

Simon HA. 2002. Near decomposability and the speedof evolution. Industrial and Corporate Change 11(3):587–599.

Somaya D, Teece DJ. 2007. Patents, licensing andentrepreneurship: effectuating innovation in multi-invention contexts. In Entrepreneurship, Innovation,and the Growth Mechanism of the Free-Market


D. J. Teece

Enterprise, Sheshinski E, Strom RJ, Baumol WJ(eds). Princeton University Press: Princeton, NJ;185–212.

Teece DJ. 1976. Vertical Integration and VerticalDivestiture in the U.S. Oil Industry . StanfordUniversity Institute for Energy Studies: Stanford, CA.

Teece DJ. 1977. Technology transfer by multinationalenterprises: the resource cost of transferring tech-nological know-how. Economic Journal 87: (June):242–261.

Teece, DJ, 1980. Economies of scope and the scopeof the enterprise. Journal of Economic Behavior andOrganization 1(3): 223–247.

Teece DJ. 1981. Internal organization and economic per-formance: an empirical analysis of the profitability ofprincipal enterprises. Journal of Industrial Economics30(2): 173–199.

Teece DJ. 1982. Towards an economic theory of themultiproduct firm. Journal of Economic Behavior andOrganization 3(1): 39–63.

Teece DJ. 1986. Profiting from technological innovation.Research Policy 15(6): 285–305.

Teece DJ. 1988. Technological change and the natureof the enterprise. In Technical Change and Eco-nomic Theory , Dosi G, Freeman C, Nelson RR, Sil-verberg G, Soete, L (eds). Pinter: London; 256–281.

Teece DJ. 1990. Contributions and impediments of eco-nomic analysis to the study of strategic management.In Perspectives on Strategic Management , Fredrick-son JW (ed.). HarperCollins: New York; 39–80.

Teece DJ. 2000. Managing Intellectual Capital: Orga-nizational, Strategic, and Policy Dimensions . OxfordUniversity Press: Oxford, U.K.

Teece DJ. 2003. Expert talent and the design of(professional services) enterprises. Industrial andCorporate Change 12(4): 895–916.

Teece DJ. 2006. Reflections on profiting from innovation.Research Policy 35(8): 1131–1146.

Teece DJ. 2007. Managers, markets, and dynamic capa-bilities. In Dynamic Capabilities: UnderstandingStrategic Change in Organizations , Helfat C, Finkel-stein S, Mitchell W, Peteraf MA, Singh H, Teece DJ,Winter SG (eds). Blackwell: Oxford, U.K.; 19–29.

Teece DJ, Pisano G. 1994. The dynamic capabilities ofenterprises: an introduction. Industrial and CorporateChange 3(3): 537–556.

Teece DJ, Pisano G, Shuen A. 1990a. Enterprise capabil-ities, resources and the concept of strategy. Consor-tium on Competitiveness and Cooperation, Workingpaper CCC 90-8, Institute of Management, Innovationand Organization, University of California, Berkeley,CA.

Teece DJ, Pisano G, Shuen A. 1990b. Firm capabilities,resources and the concept of strategy. EconomicAnalysis and Policy Working Paper EAP-38,Institute of Management, Innovation and Organization,University of California, Berkeley, CA.

Teece DJ, Pisano, G, Shuen, A. 1997. Dynamic capabili-ties and strategic management. Strategic ManagementJournal 18(7): 509–533.

Tushman M, Anderson P. 1986. Technological disconti-nuities and organizational environments. Administra-tion Science Quarterly 31: 439–465.

Utterback J, Suarez F. 1993. Innovation, competition, andmarket structure. Research Policy 22(1): 1–21.

Wernerfelt B. 1984. A resource-based view of the firm.Strategic Management Journal 5(2): 171–180.

Williamson OE. 1975. Markets and Hierarchies . FreePress: New York.

Winter SG. 2003. Understanding dynamic capabilities.Strategic Management Journal , October Special Issue24: 991–996.

Zott C. 2003. Dynamic capabilities and the emergenceof intraindustry differential firm performance: insightsfrom a simulation study. Strategic ManagementJournal 24(2): 97–125.


Documents

Explicating dynamic capabilities: the nature and ...pages.stern.nyu.edu/~eyoon/seminar/CrossDisciplinary/...Explicating Dynamic Capabilities: Nature and Microfoundations must the innovating