RealOptionsandITPlatformAdoption: …gkmc.utah.edu/7910F/papers/Real Options and IT Platform... · 2016. 5. 23. · Information Systems Research Vol.15,No.2,June2004,pp.132–154

Information Systems ResearchVol. 15, No. 2, June 2004, pp. 132–154issn 1047-7047 �eissn 1526-5536 �04 �1502 �0132

informs ®

doi 10.1287/isre.1040.0021©2004 INFORMS

Real Options and IT Platform Adoption:Implications for Theory and Practice

Robert G. FichmanWallace E. Carroll School of Management, Boston College, 452B Fulton Hall, 140 Commonwealth Avenue,

Chestnut Hill, Massachusetts 02467-3808, [email protected]

The decision processes surrounding investments in innovative information technology (IT) platforms arecomplicated by uncertainty about expected payoffs and irreversibilities in the costs of implementation. Whenuncertainty and irreversibility are high, concepts from real options should be used to properly structure theevaluation and management of investment opportunities, and thereby capture the value of managerial flexibility.However, while innovation researchers have posited that option value can influence the motivations of earlyadopters, and options researchers have identified emerging IT as a promising area for application of optionsvaluation techniques, there has yet to be a systematic theoretical integration of work on IT innovation and realoptions.This paper seeks to fill this gap by developing a model of the determinants of option value associated with

investments in innovative IT platforms. In so doing, the model addresses a central question in the innovationfield: When should a firm take a lead role in innovation with emerging technologies? The analysis begins withan explanation of real options analysis and how it differs from conventional approaches for evaluating newtechnologies. Then a set of 12 factors—drawn from 4 complementary perspectives on organizational innovation(technology strategy, organizational learning, innovation bandwagons, and technology adaptation)—is synthe-sized into a model of the option value of IT platform investments. Rationales are provided to explain the directeffects of these factors on option value, and selected interactions among the factors are also considered. Finally,the implications of the model are presented in three areas: predicting IT platform initiation and adoption, valu-ing IT platform options, and managing IT platform implementation.

Key words : real options; IT investments; IT platforms; IT adoption; IT innovationHistory : Robert W. Zmud, Senior Editor. This paper was received on August 29, 2002, and was with theauthor 11 months for 3 revisions.

1. IntroductionThe pace of change in the information technology (IT)field has been rapid over the past decade, with a hostof promising new platform technologies1 confrontingforward-looking organizations. Managers know theymust innovate (at least occasionally) to thrive, yetit can be difficult to decide which technologies toadopt, when to adopt them, and how to managethe implementation process to realize business value.

1 An IT platform is broadly defined here as a general-purpose tech-nology that enables a family of applications and related businessopportunities. This includes computing platforms (e.g., Palm OS),infrastructure platforms (e.g., wireless networking), software devel-opment platforms (e.g., Java), and enterprise application platforms(e.g., ERP). Thus, the term “IT platform” may be viewed as a gen-eralization of the term “software platform” employed by Taudeset al. (2000).

Much of this difficulty arises from two challenges typ-ically associated with IT platform innovations: uncer-tainty about the benefits of using the innovationand irreversibility in the costs of deployment. Uncer-tainty arises both from the unpredictable evolution ofthe technologies themselves and from strategic pathdependencies they impose on a firm’s future IT tra-jectory. Irreversibility arises from high learning andadaptation costs during deployment and high switch-ing costs after deployment.When uncertainty and irreversibility are high—and

when managers have flexibility concerning the timingand structure of technology adoption investments—it is fruitful to view such investments through areal options lens (Amram and Kulatilaka 1999, Dixitand Pindyck 1994, Trigeorgis 1993). According to thisview, initial investments in new IT—such as pilot

132

Fichman: Real Options and IT Platform Adoption: Implications for Theory and PracticeInformation Systems Research 15(2), pp. 132–154, © 2004 INFORMS 133

projects, prototypes, or the first phase in a multiphaseimplementation—create “growth” options2 (Taudes1998). These options confer the right, but not the obli-gation, to obtain benefits from future deploymentsof the technology, just as financial call options con-fer the right, but not the obligation, to obtain bene-fits from future ownership of traded securities. Thoseorganizations that simply defer investment in IT plat-form technologies may not have quite the same claimon future benefits because of time-compression disec-onomies (Dierickx and Cool 1989).Viewing technology adoption as an option has a

number of implications for IT innovation. One is thatorganizations will approach the justification processrather differently. Managers will not be satisfied withthe common practice of developing a static best esti-mate of the costs and benefits associated with tech-nology deployment. Instead, they will seek to employmore dynamic valuation methods that properly reflectthe value of managerial flexibility in project execution.This could mean estimating a formal options pricingmodel (OPM), though it is not essential to use suchmodels to bring options thinking into the justifica-tion process. Managers can also account for manage-rial flexibility using decision trees (Hamilton 2000),qualitative scoring models (McGrath and MacMil-lian 2000), or even just better-informed managerialintuition. The key point is to realize that whenuncertainty and irreversibility are high, omitting thevalue of managerial flexibility can lead to substan-tial understatement of the value of investments innew IT.Beyond project justification, options thinking has

implications for how investments are managed. Firmsshould be more inclined to initiate uncertain invest-ments in IT because the option value estimate ofan uncertain investment always exceeds the static

2 Growth options refer to a situation where early investment is a “pre-requisite or a link in a chain of interrelated projects, opening upfuture growth opportunities” (Trigeorgis 1993, p. 204). Other kindsof real options are applicable to IT investments, such as optionsto defer, to change scale, to abandon, or to switch use. For exam-ple, Benaroch and Kauffman (1999b) examine the option to deferin the case of Yankee 24’s decision on timing of an investment toexpand its point-of-sale debit card network. However, this paperwill focus on the growth options associated with early investmentsin IT platforms.

net present value (NPV) estimate (and sometimes byquite a wide margin, Taudes et al. 2000). However,such firms should also be more likely to redirector terminate uncertain projects, because the optionsapproach assumes that managers can and will “cull”projects when the ball of uncertainty bounces thewrong way. This contrasts with the often observedpropensity of managers to escalate commitment totroubled projects (Keil et al. 2000).There has been a growing stream of research on IT

investments and real options (Benaroch 2002; Bena-roch and Kauffman 1999a, 1999b; Clemons 1991; DosSantos 1991; Kambil et al. 1993; Taudes 1998; Taudeset al. 2000). In addition, innovation researchers, whohave traditionally been interested in how organiza-tional factors (e.g., resources and capabilities) canaffect technology adoption, have observed the generallink between technology options and the motivationsof early adopters (Cohen and Levinthal 1990, Fich-man and Kemerer 1999). Nevertheless, there has beenno attempt to date to synthesize work on real optionsand the determinants of technology adoption, eventhough (as just argued) options represent a superiorlogic for assessing the value associated with technol-ogy adoption. Consequently, the goal of this paper isto develop a theoretical model of the determinants ofoption value associated with investments in IT plat-form innovations.The proposed model is intended to support a com-

prehensive examination of a central question in theinnovation field: When should a firm take the leadas an innovator with emerging technologies? Whilesome prior research has noted that organizations onthe leading edge in adopting highly uncertain newtechnologies often do so out of (at least implicit)recognition of option value (Cohen and Levinthal1990), this insight has not been incorporated in anysystematic fashion into prior models of innovationinitiation or adoption. The posited model is consis-tent with much prior work on the question of whichfirms should lead in innovating, but it also drawsattention to some new factors and interaction effectsnot previously considered in IT innovation research.It also suggests different relationships for some pre-viously considered factors. Although the model canbe generalized to other technologies, the focus hereis on IT platforms because, as will be argued below,

Fichman: Real Options and IT Platform Adoption: Implications for Theory and Practice134 Information Systems Research 15(2), pp. 132–154, © 2004 INFORMS

the logic of real options is particularly appropriate forsuch technologies.The remainder of this paper is structured as fol-

lows. It begins by summarizing the general case forviewing technological innovation in general, and ITplatform investments in particular, through the realoptions lens. Then a set of 12 factors—drawn from4 complementary perspectives on innovation (tech-nology strategy, organizational learning, innovationbandwagons, and technology adaptation)—is synthe-sized into a model of the determinants of option valueof innovative IT platform investments. Rationales areprovided to explain the direct effects of these factorson option value, and selected interactions among thefactors are also considered. Finally, the implicationsof this work for research and practice are offered inthree areas: models for predicting early IT platforminitiation and adoption, estimating the option value ofIT platform investments, and managing IT platformprojects.

2. Early Investments in IT PlatformInnovations

2.1. Real Options AnalysisFor many years management scholars have arguedthat we should view uncertain investments innew technology through a real options lens (Dixitand Pindyck 1994, Trigeorgis 1993, p. 204). Theseresearchers have noted the similarity between optionson physical assets and the kinds of options created bytechnology positioning investments (McGrath 1997).A technology positioning investment is an initial expen-diture on a technology (e.g., for R&D) that creates theright, but not the obligation, to obtain the benefitsassociated with further development and deploymentof the technology. This is similar to the structure of areal option, which confers the right, but not the obli-gation, to obtain the benefits associated with somephysical asset.Firms that make such investments retain full expo-

sure to the upside potential of the technology shouldsubsequent events prove favorable to deployment,but they can limit losses to just the positioning invest-ment if future events prove unfavorable. This asym-metric exposure to gains versus losses has severalimplications, some of which are counterintuitive. The

most immediate counterintuitive implication is thatignoring the effects of this asymmetry (as occurs intraditional NPV calculations) can lead to surprisinglylarge understatements in the value of the position-ing investment, and thus inhibit innovation. Anothercounterintuitive implication is that the value of a posi-tioning project will increase as the variance of poten-tial net benefits from full deployment increase, evenif the expected value of net benefits are held con-stant. That is, greater uncertainty at the outset of aninvestment increases the value of an option. Ordinar-ily, higher uncertainty, if anything, will tend to lowerthe estimated value of an investment by encouraginguse of a higher discount rate.Three conditions are prerequisite to using real

options concepts to structure the evaluation andmanagement of technology investments: uncertainlyregarding net payoffs, irreversibility in project costs,and managerial flexibility regarding how projects arestructured (Dixit and Pindyck 1994). All three con-ditions hold strongly for innovative IT platforms.Net payoffs are typically quite uncertain becauseof a combination of three characteristics that oftenattend such platforms: high interpretive flexibility(which multiplies the array of feasible implemen-tation configurations and associated values), highknowledge barriers (which make it hard to anticipatethe full costs of implementation), and strong positivenetwork externalities (which condition the value ofimplementation on whether a robust adoption net-work arises). (The role of these three characteristicsin real options analysis is considered in greater detailin §3.)Regarding the second precondition, the adoption of

an innovative IT platform is essentially an investmentin a new organizational capability, and such invest-ments are largely irreversible due to the tight couplingof technology and organization (Kogut and Kulatilaka2001). While a portion of expenditures for hardwareand software can be reversed in some cases, otherdirect costs associated with organizational learningand adaptation cannot be reversed. These costs, whichtypically dwarf the out-of-pocket costs of the tech-nology per se, include expenditures for training, hir-ing experienced workers and consultants, engaging inlearning by doing, developing new policies and pro-cedures, establishing supporting infrastructure, and


absorbing losses in productivity during the transitionfrom old to new.Finally, managers have considerable flexibility in

how they approach IT platform investments. Thisflexibility can take two basic forms: flexibility in theprocess of delivering the new system, and flexibility inthe result, i.e., what the system offers for future usesand enhancements. Flexibility in the former is pro-moted by managerial discretion in how projects aredecomposed and staged, while flexibility in the latteris promoted by interpretive flexibility as well as byproactive steps to make systems more generic, multi-purpose, interoperable, and scalable.

2.2. IT Platform Adoption and Real OptionsThe options associated with new IT platforms havesimilarities to those created by R&D (Schwartz andZozaya-Gorostiza 2003) and can be seen as a specificinstance of the general case of technology position-ing investments. The innovation process begins withsome positioning investment in the platform, whichcan take the form of a pilot project, prototype, estab-lishment of necessary infrastructure, or some base-line implementation of the platform itself. This initialinvestment positions the firm to conduct follow-onprojects associated with the platform. For example,in the ERP case examined by Taudes et al. (2000),the positioning investment was the baseline imple-mentation of R/3 from SAP, which opened the doorto follow-on projects related to EDI, workflows, ande-commerce. Based on the Black-Scholes OPM, theestimated option value of follow-on projects exceededthe conventional NPV estimates by a factor offour.While a formal OPM is not needed to apply options

thinking, it is often worthwhile to employ suchmodels, especially on major projects or in situationswith competing investment scenarios.3 Beyond the

3 The Black-Scholes OPM calculates option value based on (1) theprojected value of the project (assumed to be distributed lognor-mal and nonnegative), (2) the variability of projected value, (3) thecost of the project (assumed to be known with certainty), (4) therisk-free rate of return, and (5) the time until expiration. OtherOPMs have less-stringent assumptions. The multiplicative formof the Cox-Rubenstein Binomial Model allows nonconstant vari-ance, and the additive form of the model also allows project valueto be normally distributed with the potential for negative values

example described above, OPMs have been appliedto investments in decision support systems (Kumar1999), telecommunications IT infrastructure (Panayiand Trigeorgis 1998), ATM banking network infras-tructure (Benaroch and Kauffman 1999b), and object-oriented middleware (Dai et al. 1999).When using real options to frame the IT platform

investment problem, the cost of the initial positioninginvestment can be seen as the “price” paid to obtainthe set of options enabled by the positioning invest-ment. (A deferral option is a special case where thepositioning investment cost is zero.) Each follow-onproject enabled by the positioning investment is mod-eled as a separate option. Positioning investments cantake two basic forms: provisional adoption initiatives(prototypes or pilot projects) that allow a detailedevaluation of the technology and need not providelasting benefit in themselves, and larger “baseline”implementations of the full platform. Either way, thetotal option value of the positioning investment isequal to the sum of the option values of follow-onprojects plus the NPV (possibly negative) of the posi-tioning investment.4 While practical challenges existin the application of OPMs to IT platform invest-ments, in many cases these challenges will be man-ageable (Copeland and Tufano 2004), and where theyare not, managers can use other techniques to supportoptions thinking, such as decision tree analysis, quali-tative scoring models, or general project managementheuristics.Regardless of the approach used to value real

options—and there are many—all are consistent with

(Copeland and Antikarov 2001). Schwartz and Zozaya-Gorostiza(2003) have developed an OPM that explicitly models uncertaintyin project costs and allows an additional form of uncertainty, whichis the possibility of a catastrophic event during development. Inaddition, Monte Carlo simulation methods can be used that per-mit a great deal of flexibility in modeling options (Amram andKulatilaka 1999, Chapter 8). An excellent summary of OPMs andtheir use on IT investments can be found in Schwartz and Zozaya-Gorostiza (2003).4 For pilots and prototypes, the cost of the positioning investmentmay be modest, but because organizations have limited resources todevote to provisional adoption, the decision to initiate investmentin one technology carries the opportunity cost of not investing insome other technology that may have a much higher option value.Therefore an evaluation of the full option value enabled by a posi-tioning investment may be warranted even when the cost of thepositioning investment is comparatively low.


several stylized facts about any particular investmentscenario, as presented below. For each stylized fact acomparison is drawn with the traditional discountedcash flow (DCF) approach (provided in parentheses).To support the presentation of these stylized facts,Figures 1a and 1b provide a graphic illustration ofa hypothetical investment scenario viewed from boththe DCF and options perspectives.The stylized facts are as follows:(1) There is a wide variance of potential net payoffs

resulting from the follow-on project. This follows directlyfrom an assumption of high uncertainty in benefitsthat underlies the real options approach. Uncertaintyin costs, which may or may not be modeled depend-ing on the OPM used, will also generally increase thevariance of net payoffs.5 (This stylized fact is con-sistent with the traditional DCF view; although inpractice managers will usually estimate one deter-ministic “best guess” scenario when employing DCFanalysis.)(2) Exposure to positive and negative payoffs is asym-

metric. This follows directly from the assumption ofmanagerial flexibility underlying real options analy-sis. In particular, it is assumed that managers willhave the discretion to refuse to exercise options thatare not “in the money,” i.e., for which the updatedestimates of implementation costs exceed those forthe benefits. To put this graphically, it is assumedthat managers are only exposed to the payoff regionto the right of zero in Figure 1b. (This stylizedfact contradicts the DCF view, which assumes allinitiated projects are brought to completion, i.e.,firms are exposed to the whole payoff region inFigure 1a.)

5 Kumar (1996) has shown that when costs are allowed to vary, anincrease in the variance in either benefits or costs can theoreticallydecrease the overall variance of net payoffs. However, this can onlyoccur when changes to benefits and costs are assumed to be posi-tively correlated. While a positive correlation may seem plausible atfirst glance, a closer examination reveals that it is unlikely to occurin practice. When one adds the necessary assumption that the scopeof a follow-on project is fixed, then there is no reason to assumethat upward revisions of costs for a project should be accompaniedby upward revisions of estimated benefits. In fact, if anything, onemight expect a negative correlation because certain adverse eventscould simultaneously increase costs and lower benefits (e.g., theadopted platform loses a standards war).

Figure 1 Conventional vs. Real Options Approach (a) ConventionalDCF Approach; (b) Real Options Approach

0.09

0.02

0.180.20

0.17

0.10

0.080.07

0.050.04

The full system costs $10 m toimplement and has potentialbenefits ranging from $0 to 20 m witha resulting PDF of net payoffs below.The expected value of immediate fullinvestment is calculated bymultiplying the probability of the truepayoff falling within each 1 of the 10probability bars by the averagepayoff for the bar, i.e., 0.02*(– 9) +0.09*(–7 ) … 0.04*(+9)= negative $1.1 m.

–9 m –7 m –5 m –3 m –1 m +1 m +3 m + 5 m +7 m +$9 m

++

0.10

0.080.07

0.050.04

The system is structured as afollow-on project to be completedlater (e.g., after a pilot project). Ifit is assumed managers will knowthe true payoff at this later time,then the negative net payoffregion will be avoided. This leadsto a revised estimated payoff of0.10*(1)+ 0.08 (3) …+0.04 (9)

= $1.4 m. This is only an approximation of the value that would be estimated by a formal OPM.

+1 m +3 m +5 m +7 m $9 m

+ **

(a)

(b)

(3) The only way to avoid the negative payoff region isto cancel or redirect the follow-on project. This followsdirectly from the assumption of irreversibility under-lying real options analysis, which states that the costof investment cannot be reversed once incurred. (Thisstylized fact is not relevant to the traditional DCF


view, which assumes initiated projects are always fol-lowed to completion.)(4) The value of the option increases with increas-

ing expected value of potential payoffs. An increase inexpected value (i.e., a rightward shift of the wholeprobability density function (PDF) of payoffs in Fig-ures 1a and 1b) increases both the probability of exer-cise and the expected value if exercised. Because thevalue of an option value is proportional to the prob-ability of the option being exercised times the valueif exercised (see Figure 1b), an increase in expectedvalue results in an increase in the option value. (Thisis consistent with the DCF view, where a rightwardshift increases the value estimate.)(5) The value of the option increases with increasing

variance of potential payoffs. An increase in payoff vari-ance (i.e., a fattening of the tails of the PDF in Figure1), other things being equal, increases the expectedvalue if exercised, which, as previously stated in (4),increases the value of the option. (This contradictsthe DCF view in which increasing variance of poten-tial payoffs either has no value or may be penalizedthrough use of a higher discount rate.)(6) Increases in managerial flexibility increase the value

of an option. Increases in flexibility ensure that anyparticular option that exists in principle will be “wellformed” in practice—which is taken here to meanthat projects can be structured such that managerswill find it technically and organizationally feasibleto redirect projects or to let options expire as theynormatively should. Looking beyond any individualoption, increased managerial flexibility will increasethe array of potentially valuable options that may becreated by any given positioning investment. (Thiscontradicts the DCF view in that flexibility to termi-nate or redirect a project does not affect the estimatedvaluation.)The final three stylized facts are especially notable

in that they can be used to identify which determi-nants of the positioning investments will be espe-cially salient from an options perspective. They alsoidentify the different mechanisms by which thosedeterminants affect option value. On the first point,stylized facts (5) and (6) suggest that determinantsthat increase uncertainty and/or increase managerialflexibility will be especially salient, because uncer-tainty and flexibility affect the traditional and real

options estimates of value in different ways (as justdescribed). On the second point, stylized facts (4), (5),and (6) suggest that any determinant that increasesthe expected value of potential payoffs, increases thevariance of payoffs, or increases in managerial flexi-bility will increase the real options estimate of value.While increases in expected value and variance serveto increase the value of a “well-formed” option, man-agerial flexibility impoves the odds that a particularoption will be “well formed” and also enlarges the setof potentially valuable options.

3. A Model of Option Value inIT Platform Investments

This section presents a model of the determinants ofoption value in innovative IT platform investments.It is argued that firms that recognize the potentialoption value for IT platforms will be more likely toinitiate adoption of those platforms through technol-ogy positioning investments. Thus, the model createsa synthesis of work on real options and technologyadoption.To identify potential determinants of option value,

I considered four complimentary perspectives onorganizational innovation, labeled here as (1) tech-nology strategy, (2) organizational learning, (3) inno-vation bandwagons, and (4) technology adaptation.These perspectives, while not exhaustive,6 span alarge portion of the relevant literature. Within eachperspective I sought to identify factors that have anespecially significant effect on the variance of returnsor managerial flexibility—and thus (as explained inthe previous section) may affect option estimatesof value differently from traditional assessments ofvalue. The resulting model incorporates a dozenbroad determinants, as depicted in Figure 2 and

6 Two notable perspectives that were excluded are the institutionalview of innovation (e.g., Abrahamson 1996, DiMaggio and Powell1983) and the large stream of research on the generic properties oforganizations that make them more or less innovative (size, formal-ization, centralization, etc.). The former perspective was excludedbecause the focus here is on economic drivers of innovation. AsMcGrath et al. note, “The options lens embeds a logic for antic-ipating whether [firm behavior] makes economic sense or not”(2004, p. 98). The latter perspective was excluded because the modelseeks to understand the drivers for adoption of a particular ITplatform.


Figure 2 Antecedents of Option Value in IT Platform Investments

Technology strategy perspective

• Radicalness

• Strategic importance of affected products or processes

• Sustainability of advantage

• Innovative capabilities/endowments

Optionvalue

Organizational learning perspective

• Knowledge barriers

• Learning-related endowments

• Contributions to exploitable absorptive capacity

Bandwagon perspective

• Susceptibility to network externalities

• Prospects for network dominance—class

• Prospects for network dominance—instance

Adaptation perspective

• Interpretive flexibility

• Divisibility

IT platformpositioninginvestmentssupportingadoption

Table 1. Given the historical fragmentation of researchon organizational innovation, there is value in syn-thesizing elements from four important streams intoa unified model that applies to a large class of infor-mation technologies.However, the proposed model is not just a super-

set of factors previously considered in prior innova-tion research. Rather, it emphasizes factors that havea clear economic interpretation from a real optionsperspective and excludes other factors that do not.7

Option value is placed as a mediating variable lead-ing to adoption because managers are posited to takeinto account the economic value of initial adoption—which is best captured by the real options valuation

7 A few prominent examples of excluded variables are thoseassociated with the management team (top management sup-port, presence of a champion), communication-related variables(e.g., investments in communication channels), and organizational-structure-related variables (e.g., centralization, formalization).

logic—when making investment decisions. This leadsto an initial proposition:

Proposition 0. Increased option value will increasethe propensity of firms to make positioning investmentsthat support the initial adoption of IT platforms.

Prior studies within the four perspectives vary inhow explicitly they consider economic considerationsin theorizing. Rationales will often be articulated interms of the effects of determinants on costs and ben-efits or the probability of success. The implication,often left unstated, is that some adopters will havehigher expected returns from adoption than others.Such rationales are therefore consistent with the DCFor informal cost-benefit approaches to valuation com-monly employed. In the proposed model, the eco-nomic logic for these variables is made explicit andexpanded to include options. The option value logicoverlaps with the implicit DCF logic of prior innova-tion models when considering the expected value of


Table 1 Determinants of Option Value

Determinant Intermediate mechanisms Overall effect

EV of Var. of Option value DCF valueLabel Definition payoffs payoffs Flexibility estimate∗ estimate

1. Radicalness The extent of potential improvements in organiza-tional products or processes enabled by the tech-nology

+ +

2. Strategic importance ofaffected products orprocesses

The extent to which products or processes potentiallyimproved by the innovation are central to the com-petitive position or value proposition of the firm

+ +

3. Sustainability of competitiveadvantage

The extent to which expected improvements to afirm’s strategically important products or prod-ucts resist rapid duplication by competitors

+ + ++ +

4. Innovative capabilitiesand endowments

The extent to which an organization possessesresources (human, technical, organizational) con-ducive to effective deployment of the innovation

+ + ++ +

5. Knowledge barriers The extent of the burden of organizational learningassociated with adoption

− + ? −

6. Learning-related endowments The extent to which an organization possessesknowledge, skills, routines, incentives, and otherresources conducive to effective organizationallearning surrounding the innovation

+ + ++ +

7. Contributions to exploitableabsorptive capacity

The extent to which knowledge to be gained duringdeployment contributes to absorptive capacity indomains with long-lasting strategic relevance

+ +

8. Susceptibility to networkexternalities

The extent to which a technology increases in value toindividual adopters with the size of the adoptionnetwork

+ + ++ +

9. Prospects for networkdominance of thetechnology class

The extent to which the innovation’s technology classis likely to achieve a dominant position relative tocompeting technology classes

+ + +

10. Prospects for networkdominance of the technologyinstance

The extent to which the technology instance beingadopted is likely to achieve a dominant positionrelative to competing technology instances withinthe same class

+ + +

11. Interpretive flexibility The extent to which a technology permits multipleinterpretations on the part of adopters about howit should be implemented and used

+ + + +++ +

12. Divisibility The extent to which a technology can be divided forsequential implementation in such a way that eachincremental segment positions the firm for a posi-tive payoff, even if no further implementation seg-ments are pursued

+ + ++ +

∗ The presence of multiple plus signs indicates the presence of more than one intermediate mechanism promoting higher option value.

investments, but differs when considering the contri-bution to payoffs from variability of returns and man-agerial flexibility. Thus, as will be discussed later, themagnitude of posited effects for included variables—and in a few cases, the direction of effects—is dif-ferent owing to the options perspective. Also, themodel suggests some variables and interaction effectsthat have not been previously studied in the ITcontext.

Of the 12 determinants in the model, 7 (rad-icalness, knowledge barriers, susceptibility to net-work externalities, prospects for dominance of thetechnology class and instance, interpretive flexibility,and divisibility) should be viewed as characteristicsof the technology context surrounding a particu-lar adoption opportunity, while the remaining fiveare characteristics of the organizational context. Thecharacteristics of the technology context can be


viewed as either invariant across organizations (ifthey are defined by how the average or typical adopt-ing organization would experience the characteris-tic8) or as varying across organizations (if definedby how a particular organization would perceive orexperience the characteristic) (see Fichman 2000 fora more detailed discussion). Thus, these factors canbe measured either at the level of the technology orat the level of the technology-organization combina-tion, depending on the objectives and design of theresearch. The organizational factors all vary, depend-ing on the specific technology in question, and thusshould be measured at the technology-organizationcombination.In the rationales presented below, a determinant

will be considered to increase option value if it tendsto increase the expected value of potential returns,increase the variance of potential returns, or increasemanagerial flexibility in the structuring/exercise ofoptions (as per stylized facts (4)–(6) presented in theprior section). Thus, the options approach is simi-lar to traditional DCF-oriented valuation heuristicsregarding factors that increase the expected value ofreturns, but differs by assigning a positive role toincreases in variance of returns and managerial flex-ibility. Table 1 provides a summary of the positedeffects of each determinant and a comparison of theoverall impact from the options perspective versusthe DCF perspective.

3.1. Technology Strategy PerspectiveA large body of literature has looked at innova-tion as a strategic move intended to build or rein-force the competitive advantages of the firm (seeAfuah 1998, Chapter 2, for a concise review of thisresearch). Traditionally, this literature has taken a

8 Prior work suggests multiple tactics for measuring characteris-tics at the level of the technology. Meyer and Goes (1988) used anexpert panel of medical college faculty to rate each of 10 innova-tions in medical technology on observability, the level of risk, andthe degree of specialized skill required. Dewar and Dutton (1986)used the average of perceptions of respondents in a survey sampleto measure radicalness of a set of six footwear industry innovationson a continuous scale. Ettlie et al. (1984) relied on their own judg-ments to classify a set of six food-packaging innovations as eitherradical or incremental.

more structural view of which features of technolo-gies and organizations are most important. Regard-ing technology characteristics, the most frequentlymade distinction is whether the innovation is rad-ical or incremental (Damanpour 1988, Dewar andDutton 1986, Ettlie et al. 1984).9 With regard to orga-nizational characteristics, research has examined therole of many variables—size, structure, culture, staffexpertise, slack resources, etc.—in facilitating or hin-dering innovation (see Damanpour 1991 and Wolfe1994 for reviews).More recently, the technology strategy perspective

has been increasingly guided by the resource-basedview of the firm, which is primarily concerned withunderstanding the conditions that lead to sustainedcompetitive advantage (Barney 1991, Wernerfelt1984). According to this view, a resource will con-tribute to sustained advantage only if it is valuable,heterogeneously distributed among firms, and immo-bile. Because the third criterion is the least straightfor-ward of the three, it has received the most attention.Mata et al. (1995) suggest that resources will be moreimmobile when they are tied to a firm’s unique posi-tion in history, are difficult to observe due to causalambiguity, or are difficult to observe and enact due tosocial complexity. As a result, the resource-based viewemphasizes unique resources and contextual condi-tions, rather than the more generic firm characteristicsconsidered in the prior strategy literature. Relat-edly, path dependencies resulting from prior learning,investment, and experience will give firms preferen-tial advantages in exploiting real options and therebymagnify option value (Sambamurthy et al. 2003).A review of the technology strategy literature

yielded a set of four broad factors that have espe-cially significant effects on uncertainty or flexibility,and hence are especially salient from the options per-spective: (1) the radicalness of the technology, (2)the strategic importance of affected products or pro-cesses, (3) the sustainability of advantages conferredby improvements in these products or processes, and(4) the extent to which the firm possesses innovative

9 Other distinctions have related to whether an innovation enhancesor destroys competence (Tushman and Anderson 1986), is architec-tural or modular (Henderson and Clark 1990), and is disruptive orsustaining (Bower and Christensen 1995).


capabilities and endowments consistent with adop-tion of IT platforms.

3.1.1. Radicalness. The incremental versus radi-cal distinction has long been a central one in the inno-vation field. This distinction is important in its ownright and tends to correlate with other key distinc-tions (i.e., both disruptive and competency-destroyinginnovations tend to be more radical). Radicalness canbe defined in terms of the outcome of innovation orthe process of innovation. Economists focus on out-comes and define an innovation as radical when itreduces the cost of production so far that it makesthe methods employed by incumbent firms obsolete(Henderson 1993). Foster (1986) also focuses on out-comes and defines a radical innovation as one thatimplies a switch to a new improvement S-curve.10

Consistent with these outcome-oriented approaches,radicalness is defined here as the extent of potentialimprovements in organizational products or processesenabled by the technology platform.The effects of radicalness on the expected value

of returns from innovation is unclear because thepotentially greater returns are offset by correspond-ingly greater expenses. That is, technologies thatenable more radical improvements typically requiremore substantial complementary changes to orga-nizational structures, routines, and policies. On theother hand, increased radicalness clearly magnifiesthe uncertainty of early investments in IT platformsand, correspondingly, the variance of potential pay-offs. More radical innovations tend to replace, ratherthan build on, existing technologies and can havewide-ranging effects on the structure of businessesor even entire industries, both intended and unin-tended (Henderson and Clark 1990). Innovationswith incremental impacts, by contrast, tend to bemore localized, to build on existing technologies andcompetencies, and to have better-defined potentialconsequences, all of which leads to comparativelylower variance of outcomes. The above argumentsresult in the following proposition:

10 Damanpour focuses on the innovation process, arguing that “rad-ical innovations � � �produce fundamental changes in the activities ofthe organization and represent clear departures from existing prac-tice” (1988, p. 550). Rogers (1995) and Dewar and Dutton (1986) alsoemphasize process in defining radicalness as the degree of changethrust on adopters.

Proposition 1.1. Increasing radicalness increases thevariance of potential returns and thus increases the optionvalue of positioning investments in IT platforms.

3.1.2. Strategic Importance of Affected Productsor Processes. Organizations adopt new platforms tofacilitate the delivery of new applications, which,in turn are intended to improve the productsor processes of the firm. The strategic importanceof affected products or processes is defined here asthe extent to which the products or processes poten-tially improved by the innovation are central tothe competitive position or value proposition ofthe firm. This will have a considerable impact on theoption value of a proposed investment. Which pro-cesses and products are central will vary dependingon a firm’s generic strategies (operational effective-ness, strategic positioning) and the specifics of itsproduct offerings, customer needs, and competitiveenvironment.Because, in the end, it is a firm’s degree of value

added that leads to profits (Porter 2001), it can beexpected that a platform that affects the most strategi-cally important products or processes will produce agreater opportunity for outsized economic rents thanone that primarily affects low valued added or sup-port processes. This is especially true with regardto early investments because the extra advantage ofbeing a first mover or a fast follower with respect tonew products can translate into much greater marketshare and profits flowing from the product (Kesslerand Chakrabarti 1996, Lieberman and Montgomery1988). On the other hand, when a strategic initiativegoes awry, costs can be especially heavy. For example,Federal Express lost hundreds of millions of dollarsin a failed attempt to introduce ZapMail, a new plat-form for high-quality facsimile transmittal (Kemererand Sosa 1991).As was the case with radicalness, the effect of

strategic importance on the expected value of returnsis unclear. Such initiatives tend to have a greaterpotential value but also to cost more, and there isno obvious rationale for why either effect wouldsystematically predominate. However, by magnify-ing the range of net payoffs in both the positiveand negative directions, increased strategic impor-tance can be expected to dramatically increase the


variance of potential returns, leading to the followingproposition:

Proposition 1.2. Increasing the strategic importanceof affected products or processes increases the variance ofpotential returns and thus increases the option value ofpositioning investments in IT platforms.

3.1.3. Sustainability of Competitive Advantage.When a new IT platform provides the opportunity foradvantage in strategically important areas, a closelyrelated issue is the sustainability of those advan-tages. Sustainability is defined here as the extent towhich expected improvements to a firm’s strategicallyimportant products or processes resist rapid dupli-cation by competitors. According to the resource-based view of the firm, a nonproprietary innova-tion will only produce sustainable advantage whencombined with some distinctive and difficult-to-copyassets (Barney 1991). This is especially true with ITplatform innovations because they are commerciallysupplied and thus available to all (Bharadwaj 2000,Kayworth et al. 2001, Mata et al. 1995). These distinc-tive assets can include proprietary products, strongbrands, a particularly skilled workforce, or a uniquevalue chain. For example, at Dell Computer the ben-efits flowing from investments in innovative IT havebeen sustained in part due to Dell’s uniquely config-ured value chain (Magretta 1998, Porter 2001). Like-wise, due to its leadership position in ATM infrastruc-ture in New England, Yankee 24 had an unmatchedopportunity to rapidly deploy a new line of POSdebit cards (Benaroch and Kauffman 1999b). Otherfactors that increase the sustainability of advantageassociated with product innovations and hence optionvalue include high-entry barriers and switching costs(McGrath 1997) and the potential to produce lock-outthrough positive feedback loops in adoption (Schilling1998). For example, eBay’s success in maintainingtotal dominance in online auctions in the UnitedStates, even in the face of well-financed attempts atmarket entry by leading companies like Yahoo andMicrosoft, provides a clear example of lock-out.An investment that can produce a more sustainable

advantage will be more valuable than an improve-ment that can easily be copied or otherwise matchedbecause a sustainable advantage will lead to a streamof rents with a longer duration (McGrath 1997). This

will lead to an increase in the expected value of poten-tial returns. However, high sustainability should alsoincrease the variance of returns, as it creates an open-ended stream of excess rents that will tend to persistuntil some major technological discontinuity occurs.Because the duration until the emergence of a discon-tinuity is highly unpredictable, contributions to valuefrom sustainability should also be highly variable.

Proposition 1.3. Increasing sustainability of compet-itive advantage increases the expected value and variancepotential returns and thus increases the option value ofpositioning investments in IT platforms.

3.1.4. Innovative Capabilities and Endowments.Firms vary greatly in terms of the innovation-relatedcapabilities and endowments they bring to the tablewhen considering an IT platform investment. Perti-nent capabilities and endowments include supportivesenior management, effective innovation sponsorsand champions, access to key resources (technolog-ical, financial, managerial), an innovation-orientedculture, and a staff with prior experience and suc-cess in the process of innovation itself (Rogers 1995,Tornatzky and Fleischer 1990, Wolfe 1994). Firmsthat possess such endowments can innovate moreeconomically and with greater probability of suc-cess, and this should increase the expected value ofreturns from investment. For example, Microsoft iswell known for its success in leveraging unparalleledfinancial resources, brand, and dominant positionsin key existing platforms (e.g., operating systems)to establish strong positions with respect to adja-cent emerging platforms (e.g., databases, office suites,browsers).It may also be argued that organizations possessing

such endowments will have a greater ability to recog-nize and exploit a larger array of immediate follow-on projects. For example, organizations with greaterinnovation experience may have more effective rou-tines in place for scanning and evaluating opportu-nities. An increased ability to recognize and exploitfollow-on projects serves to increase managerial flexi-bility in the structuring and executing of options asso-ciated with the positioning investment.

Proposition 1.4. Increases in innovation capabilitiesand endowments increase the expected value of potentialreturns and the level of managerial flexibility and thus


increase the option value of positioning investments in ITplatforms.

3.2. Organizational Learning PerspectiveMany scholars have noted the intimate relationshipbetween organizational learning and technologicalinnovation (Cohen and Levinthal 1990, Kogut andZander 1992, Leonard-Barton 1988a, Pennings andHarianto 1992). At any particular time an organiza-tion possesses some bundle of skills and routines(Nelson and Winter 1982), and the process of technol-ogy innovation can be seen as the means by whichan organization moves from one bundle to the next.Not surprisingly then, the organizational learningperspective has inspired several studies of IT adop-tion and assimilation (Armstrong and Sambamurthy1999, Boynton et al. 1994, Fichman and Kemerer 1997,Purvis et al. 2001).A review of the literature yielded three factors from

this domain that have especially significant effects onthe variance of expected returns or managerial flex-ibility: the extent of knowledge barriers imposed bythe platform, the degree of learning-related endow-ments enjoyed by the organization, and the extent towhich adoption increases the exploitable absorptivecapacity of the firm.

3.2.1. Knowledge Barriers. In a seminal article,Attewell (1992) introduced the concept of knowledgebarriers, defined here as the extent of the burden oforganizational learning associated with adoption.11

Attewell argued that the adoption of complex orga-nizational technologies should be viewed as a specialcategory of innovation because of the special burdenof organizational learning they impose on adopters.He notes that the know-how and technical knowl-edge associated with such technologies is tacit andrelatively immobile, and has to be recreated by usersvia the processes of learning by doing (among pro-ducers) and learning by using (among adopters). Oth-ers have made similar distinctions between classes of

11 Some researchers have virtually equated radicalness with theamount of learning required (Kogut and Kulatilaka 2001, Rogers1995). However, for this analysis I treat knowledge barriers as con-ceptually distinct. It is quite possible for a technology to be diffi-cult to understand and use, but to still have incremental (i.e., fairlymodest and localized) effects on performance. Preserving this dis-tinction allows insights not immediately apparent from the prop-erty of radicalness as I have defined it.

knowledge. Von Hippel (1994) asserts that the knowl-edge used in technical problem solving is “sticky,”i.e., costly to acquire, transfer, and use in a newlocation. Badaracco (1991) distinguishes “migratoryknowledge,” which can be transferred via books,formulas, and machines, from “embedded knowl-edge,” such as individual craftsmanship, know-how,and team-based knowledge. This distinction betweenclasses of knowledge is crucial: If the knowledgeneeded to use complex technologies were not “tacit,”“sticky,” and “embedded,” then it could be readilybundled with the artifacts embodying the innovationand incorporated into the purchase price.Other things being equal, higher knowledge barri-

ers should substantially increase the costs of follow-on projects enabled by positioning investments, andthis serves to lower the expected value of poten-tial returns. In addition, because of high knowl-edge barriers, early adopters may find it difficultto judge whether successful implementation is wellwithin their organizational capabilities, or exceedsthose capabilities. This increases the chance of majorimplementation fiascoes and ensuing large opera-tional losses, such as those experienced by FoxMeyer(Bulkeley 1996) and Hershey Foods (Stedman 1999)in the wake of ill-fated ERP implementation projects.This, too, should lower the expected value of returns.On the other hand, knowledge barriers, by their

very nature, make it difficult to anticipate the fullcosts and benefits of IT platform adoption, and thismagnifies uncertainty about the net payoffs flow-ing from early investment. This serves to increasethe variance of potential returns. Thus knowledgebarriers, by lowering the expected value of returnsbut increasing their variance, have an unclear over-all effect on option value—Although on an intuitivelevel it seems likely that the former effect will usuallydominate the latter.

Proposition 2.1. Increases in knowledge barriers willdecrease the expected value of returns but will also increasethe variance of potential returns; the net effect on optionvalue will depend on which effect dominates.

3.2.2. Learning-Related Endowments. Organiza-tions have varying levels of learning-related endow-ments to bring to bear on a proposed innovation


effort. These endowments include a capable and tech-nologically up-to-date staff, a large base of applicationopportunities over which to amortize learning costs,a high degree of knowledge and skill in the vicinityof the new technology, a wide diversity of knowl-edge, and the adoption of principles consistent withlearning organizations (Armstrong and Sambamurthy1999, Fichman and Kemerer 1997, Purvis et al.2001, Swanson 1994). Learning-related endowments aredefined here as the extent to which an organizationpossesses knowledge, skills, routines, incentives, andother resources conducive to effective organizationallearning surrounding the innovation.As with innovation-related endowments discussed

above, firms that possess greater learning-relatedendowments can innovate more economically andwith a greater probability of success. More specifi-cally, increased endowments will tend to lower thecosts of platform implementations, which serves toincrease the expected value of payoffs (Fichman andKemerer 1997). High endowments (especially thosethat promote greater absorptive capacity) should alsoincrease the ability of an organization to recognizeand exploit a larger array of follow-on projects (Cohenand Levinthal 1990, Zahra and George 2002), andshould therefore serve to increase managerial flexibil-ity in the structuring and execution of options associ-ated with the positioning investment.

Proposition 2.2. Increases in learning-related endow-ments increase the expected value of potential returns andthe level of managerial flexibility and thus increase theoption value of positioning investments in IT platforms.

3.2.3. Contributions to Exploitable AbsorptiveCapacity. As Cohen and Levinthal (1990) argue intheir work on innovation and organizational learn-ing, innovative activities (such as R&D) producetwo outputs: the intended results of the effort itselfand a more indirect benefit stemming from increasesin the absorptive capacities of the firm. Schilling(1998, p. 272) expands on this argument, notingthat “through investment in technology developmentand its associated learning, firms both expand theirknowledge and skill base (or core capabilities) andimprove their ability to assimilate and utilize futureinformation (their absorptive capacity).”

The value of increased absorptive capacity dependslargely on the likelihood that future technologi-cal developments will occur in the vicinity of theabsorptive capacity being acquired today because thisincreases the managerial opportunities and associ-ated flexibility to actually exercise the options enabledby increased absorptive capacity. Thus, what mattersmost in terms of increasing managerial flexibility isthe degree of exploitable absorptive capacity, definedhere as the extent to which knowledge to be gainedduring deployment contributes to absorptive capac-ity in domains with long-lasting strategic relevance.When managers have greater flexibility in optionsrecognition and execution, this will increase optionvalue, leading to the following proposition:

Proposition 2.3. Increases in exploitable absorptivecapacity increase managerial flexibility to pursue currentlyunforeseen follow-on investments and thus increase theoption value of positioning investments in IT platforms.

Despite considerable attention to absorptive capac-ity in the management literature (Zahra and George2002), it appears that no prior work has addressedthe issue of exploitability defined here. Nevertheless,it is possible to develop a sense for when an inno-vation is part of a broad and lasting trend, eventhough the specific form of the trend is uncertain,as is the optimal timing for adoption. For example,in the early 1980s some telecommunications compa-nies began laying fiber-optic cables. Although therewas much uncertainty about whether these invest-ments were being made at the right time, there wasno doubt that the future of high-capacity transmis-sion was fiber, given the substantial inherent advan-tages over copper wire and the rapid improvementof the technology. Likewise, while there was muchdoubt in the late 1980s about the immediate payoffsfrom early investments in client server technologies(C/S) and graphical user interfaces (GUIs), it wasclear that these technologies were destined to even-tually become standard elements of modern informa-tion systems. A few years later, firms with C/S andGUI experience were better positioned to adopt objecttechnology, which diffused in a cluster of technologiesthat incorporated C/S and GUI elements (Fichmanand Kemerer 1997). When managers have confidence


that the nature of uncertainty is when an emerg-ing innovation will dominate rather than whether itwill dominate, this suggests higher exploitability ofabsorptive capacity acquired from early investment.

3.3. Technology Bandwagon PerspectiveThe diffusion patterns of many technologies followa bandwagon dynamic where adoption begets moreadoption, leading to a self-reinforcing pattern of dif-fusion. This self-reinforcing pattern can arise frominstrumental effects, such as those associated withpositive network externalities, and also from nonin-strumental effects, such as from institutional pressuresor the forces of fad and fashion (Abrahamson andRosenkopf 1993). Here I focus on the former, as theposited model is concerned with economic drivers ofinvestment in new technology.One of the key characteristics of technology band-

wagons is a tendency toward extreme diffusion out-comes. When a critical mass of adoption is reached,a sustained bandwagon and a tendency toward“winner takes all” are created (Abrahamson andRosenkopf 1997, Shapiro and Varian 1998). Alterna-tively, if a critical mass fails to develop, or if itdevelops around a different platform, the result isa “stranded” technology (Farrell and Saloner 1985).This all-or-nothing dynamic plays out at the levelof the technology and may also occur within firms(Cool et al. 1997, Markus 1987). The tendency towardextreme outcomes magnifies the uncertainty faced byany particular adopter. Thus, the bandwagon perspec-tive is particularly salient for options analysis.An examination of the literature yielded three

important factors from this perspective: susceptibil-ity to network externalities, prospects for dominanceof the emergent adoption network surrounding thetechnology class, and prospects for dominance ofthe adoption network surrounding the technologyinstance.

3.3.1. Susceptibility to Positive Network Exter-nalities. Technologies that increase in value to anyparticular adopter in proportion to the size of theadoption network possess positive network external-ities in adoption (Arthur 1988, Farrell and Saloner1987, Katz and Shapiro 1986, Schilling 1998, Shapiro

and Varian 1998).12 While all IT platforms possess net-work externalities to some extent, this varies acrosstechnologies. Thus, the more relevant issue is thedegree of susceptibility to network externalities, definedhere as the extent to which a technology increasesin value to individual adopters with the size of theadoption network.The susceptibility of an IT platform to network

externalities depends on the particular features of thetechnology and its diffusion context. More specifi-cally, this susceptibility is determined by the presenceof the following: (1) scale economies in develop-ment and production among suppliers, (2) learning bydoing among suppliers, (3) learning by using amongadopters, (4) knowledge sharing among suppliers andadopters, (5) technological interrelatedness (and asso-ciated infrastructure support), and (6) network exter-nalities arising from the exchange of information andassets conforming to standards embedded in the plat-form (Arthur 1988, Schilling 1998, Shapiro and Varian1998).When a technology is more susceptible to network

externalities, uncertainty and the resulting varianceof potential returns are magnified in two ways. First,the ultimate benefits of adoption will be determinedless by the technology as it is in its early incarna-tions and more by the technology as it will become.Early adopters are in a sense boarding a train to adestination where it is not known how far or howfast the train will go. Second, increasing susceptibil-ity to network externalities increases the probabilitythat the technology will follow the bandwagon pat-tern of extreme outcomes. This further increases thevariance of potential returns. Increased susceptibilityto network externalities also increases the value tobe derived from network benefits, and this shouldincrease the expected value of payoffs.

12 The impact of network externalities on technology adoption anddiffusion has been examined for a variety of technologies, includingspreadsheets (Brynjolfsson and Kemerer 1997), ATM networks (DosSantos and Peffers 1995), web servers (Gallaugher and Wang 2002),and electronic telephone switches (Cool et al. 1997). Althoughstrictly speaking positive network externalities represent just oneof many contributions to the more general property of increasingreturns to adoption (Arthur 1988), I will treat the term as beinginclusive of all forms of increasing returns.


Proposition 3.1. Increased susceptibility to networkexternalities increases the expected value and variance ofpotential returns and thus increases the option value ofpositioning investments in IT platforms.

3.3.2. Network Dominance of the TechnologyClass. Prospects for network dominance of the technol-ogy class is defined here as the extent to which theinnovation’s technology class is likely to achieve adominant position relative to competing technologyclasses. This contrasts with dominance pertaining tothe specific platform instance, i.e., the technologyproduct (and associated standard) being adopted (seebelow). Dominance at the level of a class mattersbecause some network benefits (benefits that flowfrom the size of the adoption network) occur at thislevel. The robust adoption of PC platforms, for exam-ple, has lowered the cost and increased the variety ofrelated components and peripherals for all platforms,not just those based on the dominant one (MicrosoftWindows). Furthermore, dominance of the technol-ogy class is required for any particular instance toachieve mass-market dominance. In evaluating theprospects for dominance of a technology class, favor-able elements include the absence of a strong installedbase for a competing technology class, strong spon-sorship, progress in standardization, and heterogene-ity of needs in the adopter population (Christensen1994, Farrell and Saloner 1985, Katz and Shapiro 1986,King et al. 1994).All classes of IT platforms possess at least some

susceptibility to network externalities. As a result, byincreasing the value of network benefits, increasedprospects for network dominance will positivelyaffect the expected value of payoffs enabled bypositioning investments. By contrast, increasing theprospects for dominance should have no systematiceffect on the variance of returns; rather, this shouldjust produce a rightward shift in the whole pay-off region. These arguments lead to the followingproposition:

Proposition 3.2. Increased prospects for dominance ofthe IT platform class increase the expected value of poten-tial returns and thus increase the option value of position-ing investments in the IT platform.

3.3.3. Network Dominance of the TechnologyInstance. Prospects for network dominance of the technol-ogy instance is defined here as the extent to which thetechnology instance being adopted is likely to achievea dominant position relative to competing technologyinstances within the same class. Even when a tech-nology class is destined for robust adoption, this doesnot ensure that a particular standard embodying thetechnology will develop a strong following. In fact,the markets for network goods are often “tippy,” withone standard eventually taking a dominant share ofthe market (Arthur 1988, Shapiro and Varian 1998).IT products are especially prone to these outcomes,as evidenced by the commanding positions of marketleaders in such areas as microprocessors, operatingsystems, office automation suites, Internet browsers,Internet routers, and databases. Being stranded witha losing technology product can be quite detrimen-tal because many network benefits accrue at the levelof the product. Also, it is usually difficult to changefrom one product to another due to product-specificinvestments and other switching costs.Shapiro and Varian (1998) have identified several

key assets that position a technology vendor to wina standards competition within a product class: con-trol over an installed base, intellectual property rights,ability to innovate, first mover advantages, manu-facturing capabilities, strength in complements, andbrand name and reputation. While this model isdirected at technology developers, it nevertheless pro-vides adopters a means to assess different technol-ogy instances. The implication is that products ratinghighly on these criteria will be more likely to win astandards competition and go on to dominance.Because all IT platform product instances possess

some susceptibility to network externalities, the valueof using the platform will be tied to the fate of thatproduct. In particular, as prospects for network dom-inance increase, the value of network benefits, and sothe overall expected value in investing in the plat-form, will likewise increase. However, as before withthe technology class, increasing prospects for domi-nance of the instance should have little or no effecton variance of payoffs. This leads to the followingproposition:

Proposition 3.3. Increases in prospects for dominanceof the IT platform instance increase the expected value of


potential returns and thus increase the option value of posi-tioning investments in the IT platform.

3.4. Technology Adaptation PerspectiveAdaptation plays a central role in the implementa-tion of advanced IT. As Tyre and Orlikowski (1993)note, “The full advantages of such technologies can-not simply be purchased off the shelf; they are wonby patiently and carefully tailoring the technology tofit a given firm’s organizational and strategic context.At the same time, organizational skills, procedures,and assumptions within the firm need to be adaptedto fit the new technology” (p. 13). Not surprisinglythen, a large body of work has examined adapta-tion processes during technology implementation anduse (DeSanctis and Poole 1994; Leonard-Barton 1988a;Orlikowski 1996, 2000; Rice and Rogers 1980; Tyre andOrlikowski 1994).While there is much diversity in this research, there

is also agreement on several key points:(1) Adaptation is possible due to the interpretative

flexibility (Orlikowski 1996) of modern IT systemsand the discretion this affords organizations to appro-priate (DeSanctis and Poole 1994) the same technol-ogy in different ways.(2) Adaptation is necessary due to “misfits”

between what the technology does “out of the box”and the current and/or desired future state of thereceiving organization (Leonard-Barton 1988a).(3) The full array of needs and opportunities to

adapt technology cannot be predicted in advancebut rather emerge during implementation itself(Orlikowski 1996).(4) The inherent unpredictability of this emergent

process suggests the need for more fluid or evenimprovisational (Orlikowski 2000) approaches toimplementation.It is somewhat ironic that while the technology

adaptation perspective has traditionally been the leastquantitative or economically minded of the four con-sidered here, its assumptions about how organiza-tions do (or should) innovate is perhaps the most wellaligned with options thinking. This perspective sug-gests two factors that are particularly salient from anoptions standpoint, in that they increase uncertaintyabout payoffs and managerial flexibility: interpretiveflexibility and divisibility.

3.4.1. Interpretive Flexibility. Interpretive flexibil-ity is defined here as the extent to which a technol-ogy permits multiple interpretations on the part ofadopters about how it should be implemented andused (Orlikowski 1996). Interpretive flexibility allowsorganizations greater discretion in how they choseto appropriate a technology and adapt it over time(DeSanctis and Poole 1994); virtually by definition thispromotes managerial flexibility in the structuring andexecution of options.The raison d’être of IT platforms is to support a wide

array of possible configurations and associated appli-cations, and so they tend to be tailorable, open-endedtools (Orlikowski 1992, 1996). However, this is moretrue for some platforms than others. An expert systemshell, for example, might be viewed as less tailorablebecause it can only be used to develop expert sys-tems, while groupware tools can be viewed as highlytailorable. Even within a technology class some prod-ucts will be more or less restrictive than others inwhat they allow adopters to accomplish (DeSanctisand Poole 1994).High interpretive flexibility serves to increase

managerial flexibility in how adoption projects arestructured by enabling a larger set of feasibleimplementation configurations. This greater set ofconfigurations also means greater uncertainty and acorresponding increase in variance in the distribu-tion of potential payoffs from adoption, because itcannot be known in advance which configurationswill be optimal. For example, Orlikowski (1996) docu-ments a groupware implementation that evolved froma tool intended to record information about help-desk incident reports into one that played important,unanticipated roles in new employee training, workerevaluations, and the distribution of work among callspecialists. By increasing the set of possible imple-mentation configurations, greater interpretive flexibil-ity should also increase the opportunity to pursuea more incremental investment strategy. As will beexplained in the following section, increased incre-mentalism tends to increase the expected value ofpotential returns and also managerial flexibility.

Proposition 4.1. Increases in interpretive flexibilityincrease the expected value of potential returns, the vari-ance of returns, and the level of managerial flexibility and


thus increase the option value of positioning investmentsin IT platforms.

3.4.2. Divisibility. A technology is divisible to theextent that it can be divided up for sequential imple-mentation in such a way that each incremental seg-ment positions the firm for a positive payoff evenif no further implementation segments are pursued(Leonard-Barton 1988b). Divisibility, then, is a keyenabler of incremental implementation,13 which liesat the heart of the options view (Bowman and Hurry1993).Increased incrementalism serves to enhance man-

agerial flexibility in options execution because atthe end of each incremental segment, managers willhave a fresh opportunity to consider which optionsare available, which should be retained, and whichshould be discarded in ensuing segments. Becausemanagers will have acquired new knowledge in situin prior segments, they will be better equipped tomake informed choices about which options to pur-sue in subsequent segments.Incremental implementation has been associated

with better outcomes in a variety of domains, includ-ing strategic decision making in high-velocity envi-ronments (Bourgeois and Eisenhardt 1988), implemen-tation of organizational innovations (Leonard-Barton1988b), and implementation of software packages(Fichman and Moses 1999). Thus, incrementalismshould also promote increases in the expected valueof returns from IT positioning investments.

Proposition 4.2. Increases in divisibility increase thepotential for incremental implementation, which in turnincreases the expected value of potential returns, the level ofmanagerial flexibility, and the corresponding option valueof positioning investments in IT platforms.

3.5. Interaction EffectsWhile space limitations do not permit a detailedexamination of interaction effects among the 12 deter-minants of option value, it seems likely that severalsuch interactions will exist. The points below are ten-tative rationales that support some of more likely

13 An incremental approach to implementation is not to be con-fused with incremental versus radical results. Radical changes canbe enacted through a sequence of incremental steps (Orlikowski1993).

interactions:• Radicalness goes to the magnitude of improve-

ments, while strategic importance goes to the lever-agability of any given improvement in terms ofproducing actual rents. Therefore, these variablesshould have a multiplicative effect. A similarly struc-tured argument can be made to posit a multiplicativerelationship between radicalness and sustainability ofcompetitive advantage.• Innovation-related capabilities and endowments

arguably become even more important for radicalinnovations because such innovations impose greaterimplementation challenges (Dewar and Dutton 1986,Ettlie et al. 1984). Thus, an interaction between rad-icalness and innovation-related endowments seemslikely. A similar argument can be used to posit aninteraction between knowledge barriers and learning-related endowments.• When network externalities are strong, net-

work benefits will comprise a higher portion of thetotal benefits associated with adoption. Because net-work benefits represent additional value beyondintrinsic benefits (which occur regardless of whetherothers adopt), higher susceptibility to network ben-efits will have a multiplicative effect on the benefitsproduced by dominance of the technology class. Asimilar argument can be made for an interaction effectwith dominance of the technology instance.• When interpretive flexibility is high, this creates

the opportunity for managers to appropriate a tech-nology in a way that is most suited to their organi-zational specifics, and this should create additionalopportunities to leverage any distinctive and hard-to-copy complementary assets that might exist. Thisshould increase the sustainability of rents flowingfrom IT platform adoption, which suggests that inter-pretive flexibility may have a multiplicative relation-ship with sustainability of competitive advantage.As these posited effects are tentative, it is suggested

that close consideration of potential interactions beconsidered in empirical work based on the proposedmodel. In addition, there may be other relationshipsamong the antecedents beyond interaction effects thatwarrant modeling. For example, it might be arguedthat interpretive flexibility and strategic importancedirectly promote greater sustainability of rents.


4. Implications and Future WorkThis article has developed the case for viewing earlyinvestments in IT platforms through the real optionslens and has identified 12 key factors that deter-mine the option value of IT platform investments.The central implications are that IT platform adop-tion is best managed as an option and that firms thatrecognize option value and manage platforms invest-ments according to the logic of real options will haveimproved returns to innovation.

4.1. IT Innovation Initiation and AdoptionThe model addresses a central question in the inno-vation field: Under what circumstances should a firmtake the lead as an innovator with emerging technolo-gies? Somewhat surprisingly, no prior attempt hasbeen made to give a comprehensive answer to thisquestion in the IT context (or complex organizationaltechnologies more generally). This question could beanalyzed using a traditional DCF logic, but adoptingan options perspective gives a richer and more accu-rate analysis, in that it both encompasses and expandson the DCF logic.The model synthesizes work from four differ-

ent streams of research (technology strategy, orga-nizational learning, innovation bandwagons, andtechnology adaptation) using a common economicperspective. In so doing, the model has highlightedsome factors—susceptibility to network externalities,interpretive flexibility, and divisibility—that have notbeen examined as innovation antecedents in priorempirical research. The model also posits severalinteraction effects that have not been previously con-sidered in IT research.14

The predictions of the model do overlap in manyareas with those suggested by a traditional DCFapproach. Eight of the factors increase the expectedvalue of projected returns, which increases the attrac-tiveness of innovation under both the traditional DCF

14 However, some early work in the innovation field employed rad-icalness as a moderator (e.g., Dewar and Dutton 1986). Also, somerecent IT studies have an embedded assumption that such moder-ating relationships exist. For example, Purvis et al. (2001) selecteda technology subject to knowledge barriers to model the impact ofvariables related to organizational learning under the presumptionthat the effects of these variables are more pertinent in the presenceof knowledge barriers.

and real options perspectives. However, most of thesefactors also increase either the variance of returns ormanagerial flexibility and so are posited to have anespecially strong impact from the options perspective.In addition, four factors have predictions that runcounter to those of the traditional view. More specifi-cally, radicalness, strategic importance, and contribu-tions to exploitable absorptive capacity are positedto have a positive relationship with adoption in theoptions model but no clear relationship with adop-tion in traditional models. Knowledge barriers, whichhave an unambiguously negative relationship in tra-ditional models, are posited to have a mixed relation-ship in the options model. This suggests that futurework could use these variables to examine whichview represents a more accurate picture of technologyadoption. For now, it is worth noting that the optionslens provides an explanation for what would other-wise seem to be a surprising lack of aversion to (oreven a preference for) more radical IT platform solu-tions. There seems to be little evidence that compar-ative radicalness slowed the rate of adoption of suchinitially popular innovations as expert systems, CASEtools, or object-oriented programming.The track record for implementation of these sorts

of IT innovations, however, is another matter. Whilethe focus here has been on adoption, it is worth paus-ing to consider the implications of real options forthe likelihood of full implementation given adoption.In both the traditional view and the options view,an increase in the expected value of returns shouldincrease the likelihood of full implementation. In theformer case, adoption is expected to always includethe full implementation, while in the latter case anincrease in expected returns increases the chance thatthe option to implement will actually be exercised.However, when the variance of returns or manage-rial flexibility is high, the situation becomes morecomplex. An increase in these variables increases thechance that an option will be created (i.e., provisionaladoption will occur) but does not increase the chancethat the option will actually be exercised. As a generalrule, organizations employing the options approachwill be more prone to initiate adoption projects butalso more prone to terminate them.15 In fact, the

15 A switch to the options perspective does not suggest an over-all increase in the incidence of implementation, though it does


greater the difference in traditional and options esti-mates of value, the less likely it is that implementa-tion will follow adoption, and so the more likely it isthat a large assimilation gap will be observed at thelevel of the population (Fichman and Kemerer 1999).Interestingly, the difference between traditional andoptions estimates of value will be particularly highwhen technologies impose high knowledge barriersand have greater susceptibility to network external-ities, and this is consistent with prior (albeit moreinformal) theorizing on the circumstances leading tolarge assimilation gaps (Fichman and Kemerer 1999).

4.2. Real Options ValuationThe model also holds implications for estimating thevalue of IT platforms. To date, leading research hasfocused primarily on developing quantitative toolsthat are most rigorous from a finance perspective.That is an admirable goal and should certainly be pur-sued going forward, but to rephrase Einstein’s apho-rism on simplicity,16 everything should be made asprecise as possible but not more so. There will be cir-cumstances where options thinking applies quite wellbut where the embedded options resist precise quan-tification, due to high levels of ambiguity and equiv-ocality surrounding the impact of these options onthe firm and its competitive environment. As Taudeset al. (2000) note in explaining managerial resistanceto traditional NPV models, it is entirely appropriatefor managers to balk at expending the effort to esti-mate a model known to be a poor representation ofreality. While the use of formal OPMs can improve ontraditional quantitative valuation models, there willstill be circumstances where any primarily quantita-tive model will require too many assumptions or sim-plifications to present an informative picture of thevalue of a project. In such circumstances, it wouldbe unfortunate if practitioners were to fall back onunguided managerial intuition rather than seek toapply the logic of real options in a systematic butqualitative fashion.The model presented here could be used to help

develop a qualitative options valuation instrument,

suggest better quality of implementation, i.e., fewer instances ofmissed opportunities or over-commitment to failing efforts.16 “Everything should be made as simple as possible, but notsimpler.”—Albert Einstein

as McGrath and MacMillian (2000) have developedfor the case of R&D and new product development.With this approach, managers assess option valueby indicating their extent of agreement or disagree-ment with several statements, each of which articu-lates some element that tends to increase or decreasethe option value of a project. For example, to evaluatethe prospects for network dominance of a technologyinstance, managers could be asked to assess (say, ona 10-point scale) whether a vendor has control overan installed base, strong intellectual property rights,strength in complements, etc. The scores on the indi-vidual statements, when aggregated, could providean initial assessment of the contribution of this onefactor to the option value of a project. A similar exer-cise could be used for the other 11 factors. While theparticulars of emerging IT platforms differ from R&Dprojects, the structure of the problem is sufficientlysimilar to encourage this approach.Even when a formal OPM is to be used, a system-

atic qualitative evaluation could still play an impor-tant role. Such an analysis could help managers tostructure a project to maximize the potential for highoption value even before the actual attempt to esti-mate the model is undertaken. (Such “amplifying”actions are further discussed in the next section.)Then, during the estimation process, the qualitativeevaluation could help managers structure the optionsframe and identify reasonable bounds on optionsmodel parameters for different options. Regarding theformer, as Amram and Kulatilaka (1999) note, it isdifficult to capture more than three or four sourcesof uncertainty in an OPM; the model introduced herecould help managers determine the most importantsources of uncertainty to capture. Regarding the lat-ter, the model helps to identify when there will bea particularly high variance of potential outcomes,which justifies relatively high values for parameterscapturing the variance of expected payoffs. Finally,this model could help managers identify the mostappropriate OPM to use. For example, when knowl-edge barriers are high and learning-related endow-ments are low, this suggests that the potential for aproject to have a negative payoff is not trivial, and soan options model that permits the value of benefits toturn negative may be warranted.


In summary, the options model introduced herecould support a systematic but qualitative evaluationof option value, which could be viewed as an endin itself or as the first step in a quantitative effort.In fact, one could imagine a triage process where aqualitative evaluation is used first to generate a roughpicture of the option value of competing initiativesand then to assess the feasibility of quantification foreach initiative. Then, when quantification is feasible,the results of the qualitative analysis could serve asone key input into the quantification process.

4.3. Project Management and CultureWhile options thinking obviously changes howinvestments are evaluated, the impacts on projectmanagement (Hamilton 2000) and culture (McGrath1999) in adopting organizations are equally far reach-ing. Organizations that evaluate investments accord-ing to the logic of real options but manage themaccording to traditional principles will be falling outof the pan of systematic undervaluation and intothe fire of systematic overvaluation. Options think-ing assumes that managers will exercise their discre-tion to reorient projects as they unfold and that thisdiscretion will be well informed and dispassionate.Therefore, firms embracing options thinking shouldbe more prone to initiate positioning investments inIT platforms, but they should also be more prone tocull options that, based on new information, are notworth exercising.As a result, options thinking will mean putting

new procedures in place and allocating resourcesto actively track the factors that affect option value.Beyond tracking, managers can also take conscioussteps to amplify the option value associated withdifferent factors (McGrath 1997). This could meanseeking out application areas with high strategicimportance and sustainability, locating innovation inareas with greater innovation- and learning-relatedendowments, biasing selection toward technologiesand products with better prospects for dominance,and pursuing more incremental implementationstrategies. In addition, options thinking means pay-ing special attention to actions that dramaticallyincrease the upside, so long as this does not have amajor adverse effect on the expected value of pay-

offs. (A further discussion of tactics for structuringIT projects to enhance option value is available inFichman et al. 2004.)Beyond implementing appropriate tools and pro-

cesses, managers must develop a culture consistentwith options thinking. Some elements of this cul-ture include a greater willingness to take on risk atthe onset of a project, an increased propensity tocritically evaluate past decisions, and—perhaps mostimportant—a commitment to resist branding termi-nated projects and the people associated with themas “failures” (McGrath 1999). As the literature on ITproject escalation shows, IT projects often take on alife of their own and can be difficult to terminate (Keilet al. 2000). Many of the qualities we prize on thepart of project team members—determination, posi-tive thinking, personal responsibility for outcomes—work against the kind of dispassionate culling ofprojects indicated by options thinking. Furthermore,some elements of options thinking itself can rein-force project escalation (Keil and Flatto 1999). Forexample, when estimated costs to complete a projecthave shifted significantly upward—but the potentialfor a very high upside potential nevertheless remainsintact—the option value of continuing can remainhighly positive despite the fact that the probability ofan overall loss on the project has increased consider-ably. As a result, options thinking will require specialattention to how managers can know when to ter-minate or redirect troubled projects. As Amram andKulatilaka (1999) observe, “In general, to capture thevalue of real options, organizations must be more flex-ible, take more risks, start a lot more projects, and killa lot of projects” (p. 210). In the end, options think-ing is just a means to promote and properly valuemanagerial flexibility in project execution; therefore, acompany culture that encourages managers to recog-nize and exploit flexibility will get the greatest benefitfrom options thinking.

5. Summary and ConclusionsThis article has developed the argument that posi-tioning investments in IT platforms (such as pilotprojects, prototypes, or the first phase in a multiphaseimplementation) create real options on the subsequentimplementation and use of the platform. Based onthis logic, a set of 12 factors—drawn from 4 comple-


mentary perspectives on organizational innovation—were synthesized into a model of the option value.By assigning an important role to factors that increasethe variance of returns and managerial flexibility, themodel presents a distinctive analysis of the determi-nants of technology adoption.The model developed here has two central implica-

tions for research and practice. First, in the domain ofIT innovation research, the model could guide workinvestigating the determinants of early initiation andadoption of IT platforms. Because increased optionvalue increases the returns to innovation, this sug-gests the factors leading to high option value shouldbe predictive of early initiation and adoption. Second,in the domain of investment valuation, the modelidentifies a set of factors that drive the option valueassociated with IT platforms. When quantitative esti-mation is feasible, an evaluation of these factors couldto help to direct managerial attention to the mostpromising options and could provide insights into thestructuring of options and the estimation of actualmodel parameters. Alternatively, when quantificationis not feasible, these factors could be used to sup-port a systematic but qualitative method for valuingIT platform options.McGrath et al. (2004) note that “Real options rea-

soning is poised to occupy a central conceptualposition in the development of theory that offersguidance for strategic decision making under uncer-tainty” (p. 86). Real options demand a differentapproach to IT platform valuation and managementof ensuing implementations, so it is essential thatresearchers and practitioners be equipped to under-stand where real options are most warranted on theo-retical grounds and which factors drive option valuein innovative IT platform investments. While optionsresearchers and innovation scholars both agree thatreal options are useful in understanding the adop-tion of emerging IT, the theoretical model developedhere represents the first synthesis of work from realoptions and IT innovation.

AcknowledgmentsThe author wishes to thank Bob Zmud, John Hogan, MarkKeil, V. Sambamurthy, Paul Tallon, the associate editor, andtwo anonymous reviewers for their many helpful commentson earlier versions of this paper.

ReferencesAbrahamson, E. 1996. Management fashion. Acad. Management Rev.

21(1) 254–285.Abrahamson, E., L. Rosenkopf. 1993. Institutional and compet-

itive bandwagons: Using mathematical modeling as a toolto explore innovation diffusion. Acad. Management Rev. 18(3)487–517.

Abrahamson, E., L. Rosenkopf. 1997. Social network effects on theextent of innovation diffusion: A computer simulation. Organ.Sci. 8(3) 289–309.

Afuah, A. 1998. Innovation Management: Strategies, Implementation,and Profits. Oxford University Press, New York.

Amram, M., N. Kulatilaka. 1999. Real Options: Managing Strate-gic Investment in an Uncertain World. Harvard Business SchoolPress, Boston, MA.

Armstrong, C. P., V. Sambamurthy. 1999. Information technologyassimilation in firms: The influence of senior leadership and ITinfrastructures. Inform. Systems Res. 10(4) 304–327.

Arthur, W. B. 1988. Competing technologies: An overview. G. Dosi,C. Freeman, R. Nelson, G. Silverberg, L. Soete, eds. TechnicalChange and Economic Theory. Pinter Publishers, London, U.K.,590–607.

Attewell, P. 1992. Technology diffusion and organizational learning:The case of business computing. Organ. Sci. 3(1) 1–19.

Badaracco, J. 1991. The Knowledge Link. Harvard Business SchoolPress, Boston, MA.

Barney, J. 1991. Firm resources and sustained competitive advan-tage. J. Management 17(1) 99–120.

Benaroch, M. 2002. Managing information technology investmentrisk: A real options perspective. J. Management Inform. Systems19(2) 43–84.

Benaroch, M., R. J. Kauffman. 1999a. A case for using real optionspricing analysis to evaluate information technology projectinvestments. Inform. Systems Res. 10(1) 70–86.

Benaroch, M., R. J. Kauffman. 1999b. Justifying electronic bank-ing network expansion using real options analysis. MIS Quart.24(2) 197–226.

Bharadwaj, A. S. 2000. A resource-based perspective on informa-tion technology capability and firm performance: An empiricalinvestigation. MIS Quart. 24(1) 169–196.

Bourgeois, L. J., K. M. Eisenhardt. 1988. Strategic decision process inhigh velocity environments: Four cases in the microcomputerindustry. Management Sci. 34(7) 816–835.

Bower, J. L., C. M. Christensen. 1995. Disruptive technologies:Catching the Wave. Harvard Bus. Rev. 73(1) 43–53.

Bowman, E. H., D. Hurry. 1993. Strategy through the optionlens: An integrated view of resource investments andthe incremental-choice process. Acad. Management Rev. 18(4)760–782.

Boynton, A. C., R. W. Zmud, G. C. Jacobs. 1994. The influence ofIT management practice on IT use in large organizations. MISQuart. 18(3) 299–318.

Brynjolfsson, E., C. F. Kemerer. 1997. Network externalities inmicrocomputer software: An econometric analysis of thespreadsheet market. Management Sci. 42(12) 1627–1647.

Bulkeley, W. M. 1996. When things go wrong.Wall Street J. (Novem-ber 18) R25.

Christensen, C. M. 1994. Exploring the limits of the technologys-curve. Prod. Oper. Management 1(4) 334–366.


Clemons, E. K. 1991. Evaluation of strategic investments in infor-mation technology. Comm. ACM 34(1) 2–36.

Cohen, W. M., D. A. Levinthal. 1990. Absorptive capacity: A newperspective on learning and innovation. Admin. Sci. Quart.35(1) 128–152.

Cool, K. O., I. Dierickx, G. Szulanski. 1997. Diffusion of innovationswithin organizations: Electronic switching in the Bell System,1971–1982. Organ. Sci. 8(5) 543–559.

Copeland, T. E., V. Antikarov. 2001. Real Options: A Practitioner’sGuide. Texere, NY.

Copeland, T. E., P. T. Keenan. 1998. How much is flexibility worth?McKinsey Quart. 2 38–49.

Copeland, T., P. Tufano. 2004. A real-world way to manage realoptions. Harvard Bus. Rev. 82(3) 90–99.

Dai, Q., R. J. Kauffman, S. T. March. 1999. Analyzing investmentsin object-oriented middleware. Workshop Inform. Tech. Systems(WITS), Charlottle, NC.

Damanpour, F. 1988. Innovation type, radicalness, and the adoptionprocess. Comm. Res. 15(5) 545–567.

Damanpour, F. 1991. Organizational innovation: A meta-analysis ofeffects of determinants and moderators. Acad. Management J.34(3) 555–590.

DeSanctis, G., M. S. Poole. 1994. Capturing the complexityin advanced technology use: Adaptive structuration theory.Organ. Sci. 5(2) 121–147.

Dewar, R. D., J. E. Dutton. 1986. The adoption of radical and incre-mental innovations: An empirical analysis. Management Sci.32(11) 1422–1433.

Dierickx, I., K. Cool. 1989. Asset stock accumulation and sus-tainability of competitive advantage. Management Sci. 35(12)1504–1514.

DiMaggio, P. J., W. W. Powell. 1983. The iron cage revisited: Institu-tional isomorphism and collective rationality in organizationalfields. Amer. Sociological Rev. 48 147–160.

Dixit, A., R. Pindyck. 1994. Investment Under Uncertainty. PrincetonUniversity Press, Princeton, NJ.

Dos Santos, B. L. 1991. Justifying investments in new informationtechnologies. J. Management Inform. Systems 7(4) 71–89.

Dos Santos, B. L., K. Peffers. 1995. Rewards to investors in inno-vative information technology applications: First movers andearly followers in ATMs. Organ. Sci. 6(3) 241–259.

Ettlie, J. E., W. P. Bridges, R. D. O’Keefe. 1984. Organization strat-egy and structural differences for radical versus incrementalinnovation. Management Sci. 30(6) 682–695.

Farrell, J., G. Saloner. 1985. Standardization, compatibility, andinnovation. RAND J. Econom. 16(1) 70–83.

Farrell, J., G. Saloner. 1987. Competition, compatibility and stan-dards: The economics of horses, penguins and lemmings. H. L.Gabel, ed. Product Standardization and Competitive Strategy. Else-vier Science, North-Holland, 940–955.

Fichman, R. G. 2000. The diffusion and assimilation of informationtechnology innovations. R. B. Zmud, ed. Framing the Domains ofIT Management: Projecting the Future Through the Past. PinnaflexEducational Resources, Inc., Cincinnati, OH.

Fichman, R. G., C. F. Kemerer. 1997. The assimilation of softwareprocess innovations: An organizational learning perspective.Management Sci. 43(10) 1345–1363.

Fichman, R. G., C. F. Kemerer. 1999. The illusory diffusion of inno-vation: An examination of assimilation gaps. Inform. SystemsRes. 10(3) 255–275.

Fichman, R. G., S. A. Moses. 1999. An incremental process for soft-ware implementation. Sloan Management Rev. 40(2) 39–52.

Fichman, R. G., M. Keil, A. Tiwana. 2004. Beyond valuation:“Options thinking” in IT project management. Working paper,Boston College, Boston, MA.

Foster, R. N. 1986. Innovation: The Attacker’s Advantage. SummitBooks, New York.

Gallaugher, J. M., Y.-M. Wang. 2002. Understanding network effectsin software markets: Evidence from web server pricing. MISQuart. 26(4) 303–327.

Hamilton, W. F. 2000. Managing real options. G. S. Day, P. J. H.Schoemaker, R. E. Gunther, eds.Wharton on Managing EmergingTechnologies. John Wiley & Sons, New York.

Henderson, R. M. 1993. Underinvestment and incompetence asresponses to radical innovation: Evidence from the Photolitho-graphic industry. RAND J. Econom. 24(2) 248–270.

Henderson, R. M., K. B. Clark. 1990. Architectural innovation: Thereconfiguration of existing a product technologies and the fail-ure of established firms. Admin. Sci. Quart. 35 9–30.

von Hippel, E. 1994. Sticky information and the locus of prob-lem solving: Implications for innovation. Management Sci. 40(4)429–439.

Kambil, A., J. C. Henderson, H. Mohsenzadeh. 1993. Strategic man-agement of information technology: An options perspective.R. D. Banker, R. J. Kauffman, M. A. Mahmood, eds. Strate-gic Information Technology Management: Perspectives on Organiza-tional Growth and Competitive Advantage. Idea Group Publishing,Middleton, PA.

Katz, M. L., C. Shapiro. 1986. Technology adoption in the presenceof network externalities. J. Political Econom. 94(4) 822–841.

Kayworth, T. R., D. Chatterjee, V. Sambamurthy. 2001. Theoreticaljustification of IT infrastructure investments. Inform. ResourceManagement J. 14(3) 5–14.

Keil, M., J. Flatto. 1999. Information systems project escalation:A reinterpretation based on options theory. Accounting, Man-agement Inform. Tech. 9 115–139.

Keil, M., J. Mann, R. Arun. 2000. Why software projects escalate:An empirical analysis and test of four theoretical models. MISQuart. 24(4) 631–664.

Kemerer, C. F., G. L. Sosa. 1991. Systems development risks instrategic information systems. Inform. Software Techn. 33(3)212–223.

Kessler, E. H., A. K. Chakrabarti. 1996. Innovation speed: A con-ceptual model of context, antecedents and outcomes. The Acad.Management Rev. 21(4) 1143–1191.

King, J. L., V. Gurbaxani, K. L. Kraemer, F. W. McFarlan, K. S.Raman, C. S. Yap. 1994. Institutional factors in informationtechnology innovation. Inform. Systems Res. 5(2) 139–169.

Kogut, B., N. Kulatilaka. 2001. Capabilities as real options. Organ.Sci. 12(6) 744–758.

Kogut, B., U. Zander. 1992. Knowledge of the firm, combinativecapabilities, and the replication of technology. Organ. Sci. 3(3)383–397.

Kumar, R. 1996. A note on project risk and option values of invest-ments in information technologies. J. Management Inform. Sys-tems 13(1) 187–193.

Kumar, R. L. 1999. Understanding DSS value: An options perspec-tive. Omega 27(3) 295–304.

Leonard-Barton, D. 1988a. Implementation as mutual adaptation oftechnology and organization. Res. Policy 17(5) 251–267.


Leonard-Barton, D. 1988b. Implementation characteristics of orga-nizational innovations. Comm. Res. 15(5) 603–631.

Lieberman, M. B., D. B. Montgomery. 1988. First mover advantages.Strategic Management J. 9 41–56.

Magretta, J. 1998. The power of virtual integration: An interviewwith Dell Computer’s Michael Dell. Harvard Bus. Rev. 76(2)72–83.

Markus, M. L. 1987. Toward a ’critical mass’ theory of interac-tive media: Universal access, interdependence and diffusion.Comm. Res. 14(5) 491–511.

Mata, F. J., W. L. Fuerst, J. B. Barney. 1995. Information technologyand sustained competitive advantage: A resource-based anal-ysis. MIS Quart. 14(3) 487–505.

McGrath, R. G. 1997. A real options logic for initiating technologypositioning investments. Acad. Management Rev. 22(4) 974–996.

McGrath, R. G. 1999. Falling forward: Real options reasoning andentrepreneurial failure. Acad. Management Rev. 24(1) 13–30.

McGrath, R. G., I. C. MacMillian. 2000. Assessing technologyprojects using real options reasoning. Res. Tech. Management43(4) 35–49.

McGrath, R. G., W. J. Ferrier, A. L. Mendelow. 2004. Real optionsas engines of choice and heterogeniety. Acad. Management Rev.29(1) 86–101.

Meyer, A. D., J. B. Goes. 1988. Organizational assimilation of inno-vations: A multilevel contextual analysis. Acad. Management J.31(4) 897–923.

Nelson, R., S. Winter. 1982. An Evolutionary Theory of EconomicChange. Harvard University Press, Cambridge, MA.

Orlikowski, W. J. 1992. The duality of technology: Rethinkingthe concept of technology in organizations. Organ. Sci. 3(3)398–427.

Orlikowski, W. J. 1993. CASE tools as organizational change: Inves-tigating incremental and radical changes in systems develop-ment. MIS Quart. 17(3) 309–340.

Orlikowski, W. J. 1996. Improvising organizational transformationover time: A situated change perspective. Organ. Sci. 7(1)63–92.

Orlikowski, W. J. 2000. Using technology and constituting struc-tures: A practice lens for studying technology in organizations.Organ. Sci. 11(4) 404–428.

Panayi, S., L. Trigeorgis. 1998. Multi-stage real options: The cases ofinformation technology infrastructure and international bankexpansion. Quart. Rev. Econom. Finance 38(4) 675–692.

Pennings, J. M., F. Harianto. 1992. Technological networking andinnovation implementation. Organ. Sci. 3(3) 356–382.

Porter, M. 2001. Strategy and the Internet. Harvard Bus. Rev. 79(3)62–78.

Purvis, R. L., V. Sambamurthy, R. W. Zmud. 2001. The assimilationof knowledge platforms in organizations: An empirical inves-tigation. Organ. Sci. 12(2) 117–135.

Rice, R., E. Rogers. 1980. Reinvention in the innovation process.Knowledge: Creation, Diffusion, Utilization 1(4) 499–514.

Rogers, E. M. 1995. Diffusion of Innovations, 4th ed. The Free Press,New York.

Sambamurthy, V., A. Bharadwaj, V. Grover. 2003. Shaping agilitythrough digital options: Reconceptualizing the role of infor-mation technology in contemporary firms. MIS Quart. 27(2)237–263.

Schilling, M. A. 1998. Technological lockout: An integrativemodel of the economic and strategic factors driving tech-nology success and failure. Acad. Management Rev. 23(2)267–284.

Schwartz, E. S., C. Zozaya-Gorostiza. 2003. Investment under uncer-tainty in information technology: Aquisition and developmentprojects. Management Sci. 49(1) 57–70.

Shapiro, C., H. R. Varian. 1998. Information Rules: A Strategic Guideto the Network Economy. Harvard Business School Press, Boston,MA.

Stedman, C. 1999. Failed ERP gamble haunts Hershey. Computer-world 33(44) 1.

Swanson, E. B. 1994. Information systems innovations among orga-nizations. Management Sci. 40(9) 1069–1092.

Taudes, A. 1998. Software growth options. J. Management Inform.Systems 15(1) 165–185.

Taudes, A., M. Feurstein, A. Mild. 2000. Options analysis ofsoftware platform decisions: A case study. MIS Quart. 24(2)227–244.

Tornatzky, L. G., M. Fleischer. 1990. The Processes of TechnologicalInnovation. Lexington Books, Lexington, MA.

Trigeorgis, L. 1993. Real options and interactions with financial flex-ibility. Financial Management 22(3) 202–224.

Tushman, M. L., P. Anderson. 1986. Technological discontinu-ities and organizational environments. Admin. Sci. Quart. 31439–465.

Tyre, M. J., W. J. Orlikowski. 1993. Exploiting opportunitiesfor technological improvement. Sloan Management Rev. 35(1)13–36.

Tyre, M. J., W. J. Orlikowski. 1994. Windows of opportunity: Tem-poral patterns of technological adaptation in organizations.Organ. Sci. 5(1) 98–118.

Wernerfelt, B. 1984. A resource-based view of the firm. StrategicManagement J. 5(2) 171–180.

Wolfe, R. A. 1994. Organizational innovation: Review, critiqueand suggested research directions. J. Management Studies 31(3)405–431.

Zahra, S. A., G. George. 2002. Absorptive capacity: A review, recon-ceptualization, and extension. Acad. Management Rev. 27(2)185–203.

RealOptionsandITPlatformAdoption: …gkmc.utah.edu/7910F/papers/Real Options and IT Platform... · 2016. 5. 23. · Information Systems Research Vol.15,No.2,June2004,pp.132–154

Documents