ABSTRACT. This paper presents a methodological approachfor the evaluation of innovation capabilities in small softwarefirms. The methodology is based on the assumption of arelationship between specific resources managed by smallsoftware firms and their innovation capabilities. Within theproposed methodological approach, a model for the quantita-tive analysis of innovation capabilities is presented. In orderto show how the methodology can be applied to concretesituations, three case studies of small firms operating in thesoftware sector and information services are presented anddiscussed.

1. Introduction

This paper is concerned with the issue of evalu-ating innovation capabilities in

small softwarefirms.

It is well known that small software firmsmust continually cope with extremely rapidchanges which demand an innovative technolog-ical and managerial response. Such a responsemust redefine the firms’ organizational assets inorder to achieve a satisfactory degree of adapta-tion to the external environment (Bellini et al.,1997). As a consequence, innovation is a neces-sary condition, not only for increasing the smallsoftware firms’ competitiveness, but primarily toensure their survival. It is an observable fact thatfirms which are not able to maintain satisfyinglevels of innovation capabilities through time

show weak performance in terms of competitive-ness and economic results (Raffa and Zollo, 1998).Thus, the problem of evaluating innovation capa-bilities is of fundamental importance both forresearchers and practitioners.

This paper tries to offer a contribution to thisissue by proposing a methodology that permits theevaluation of innovation capabilities in smallsoftware firms.

The methodology is based on the assumption ofa relationship between resources managed byfirms and their innovation capabilities.

It is well known that one of the mostimportant assumptions of the resource-based com-petition approach is that a firm’s competitiveadvantage is strictly connected to the kind and theamount of specific resources that firms are ableto acquire, develop, and manage in the course oftheir life (Amit and Shoemaker, 1993; Barney,1991; Lado and Wilson, 1994; Conner, 1991;Grant, 1991; Rumelt, 1987).

Thus, according to the resource based view,small firms’ competitiveness is linked to theircapability in acquiring and developing strategicresources. Given that for small software firmscompetitiveness equates to capability for innova-tion, our proposal is that it is possible to evaluateinnovation capabilities on the basis of the kind andthe amount of specific resources managed by smallsoftware firms during their life.

In this perspective two main methodologicalissues arise:

• How can one identify the kinds of resources onthe basis of which it is possible to measureinnovation capabilities?

• How can one measure innovation capabilitiesthrough time?

The Evaluation of Innovation Capabilities in Small Software Firms: A Methodological Approach*

Small Business Economics 21: 343–354, 2003. 2003 Kluwer Academic Publishers. Printed in the Netherlands.

Final version accepted on January 23, 2002

University of Naples Federico IIDept. of Business and Managerial EngineeringPiazzale Tecchio 80 80125 Napoli, Italy E-mail: gcapaldo@unina.it, iandoli@unina.it, raffa@unina.it, giuzollo@unina.it

Guido CapaldoLuca IandoliMario Raffa

Giuseppe Zollo

In order to determine resources linked to innova-tion capabilities, we identify a set of resources bymeans of the analysis of the literature on smallsoftware firms, as will be shown in Section 2.

Regarding the issue of evaluating innovationcapabilities through time, we refer to the life-cyclestage theory (Churcill and Lewis, 1983; Greiner,1972; Miller, 1984; Wiklund, 1996). Accordingto this theoretical viewpoint, a firm’s life can berepresented as a succession of stable phasesseparated by critical events, after which firms areforced to modify their own resources in order toadapt and react better to environmental changes.In this perspective, within each stable phase theprocess of creation and destruction of resourcescan be represented as a black-box whose inputs,represented as resources, are known.

Once that firm’s life has been described interms of stable phases and critical events, it ispossible to identify, for each phase, the resourceson the basis of which one may measure a firm’sinnovation capabilities, as will be shown in thenext section.

In particular, in the first part of the paper wedescribe the proposed methodology (Sections 2, 3,4) and in the last part it is applied to three casestudies of small software firms.

2. Evaluating small software firms’ innovationcapabilities: a resource-based perspective

As stated in the previous section, in this paperwe propose to use the resource-based theory topropose a methodology to evaluate a firm’s inno-vation capabilities. To this end, we introduce thefollowing definitions:

• the degree of market innovation capability(DMIC) as the measure of the firm’s capabilityto enhance and innovate its market in a giveninstant of time;

• the market innovation trend (LMI) as the vari-ation of the DMIC through time;

• the degree of technological innovation capa-bility (DTIC) as the measure of the firm’s capa-bility to increase its level of technologicalknow-how and expertise, in a given instant oftime;

• the technological innovation trend (LTI) as thevariation of the DTIC through time.

We propose to evaluate the degree of market inno-vation capability and the degree of technologicalinnovation capability on the basis of the followingsubset of resources:

(a) entrepreneurial resources;(b) human resources;(c) resources linked to external networks;(d) economic resources.

Each subset is a cluster of specific resources.Table I contains the list of all resources classifiedin the four clusters above.

The list of resources of Table I was obtainedthrough an analysis of the literature on innova-tive small firms (Reid and Jacobsen, 1988; Kelleyand Brooks, 1991; Quinn, 1979; Garden, 1992; DeChiara, 1996) and on software firms (Barocci etal., 1983, Raffa and Zollo, 1988, 1992, 1996). Inthe following we provide a brief description ofeach cluster:

(a) Entrepreneurial resources: this clustercontains a set of indicators concerning indi-vidual know-how and experience of the entre-preneurs. They also concern the involvementof entrepreneurs in the technical and man-agerial aspects of the firm.

(b) Human resources: this cluster includes vari-ables concerning the amount and the kind ofhuman resources available to the firms. Theimpact of human resources related to theDTIC is measured through the total number ofpeople working for the firm, the number ofsoftware developers, their level of qualifica-tion and competencies, the intensity of tech-nical training, and the presence of an intensejob-rotation policy which calls for highlyqualified resources, and allows employees tocontinually upgrade their technical know-how.The impact of human resources related to theDMIC is measured by the presence of internalemployees or external collaborators takingcare of marketing and by the intensity of mar-keting and management training activity.

(c) Resources linked to external networks: thiscluster contains variables related to the firm’sability to set up and develop contacts and col-laboration with other firms and R&D centers.Resources belonging to this cluster are usedto evaluate both the DTIC and the DMIC.

344 Guido Capaldo et al.

Frequent relationships with other firms andR&D centers focused on technical collabora-tion and an intense use of external soft-waredevelopment methodology can imply a mean-ingful increase in technological know-how.Significant relationships with other firmsderiving from commercial collaboration are anobvious indicator of a certain attention to themarket.

(d) Economic resources: information concerningprofit composition can be also useful. Highprofits in extra-regional markets, as well asa large profit deriving from commercializationof proprietary software are the result of astrong focus on the market. An intense com-

mercialization of non-proprietary hardwareand software is typical of small software firmsstrongly engaged in the development of newproducts and at the same time temporarilyunable to offer a competitive product.

As one can observe, in Table I some resources areemployed to measure both the DTIC and theDMIC at the same time; others may affect theevaluation of the DTIC and the DMIC in oppositeways. For example a deep involvement of entre-preneurs with software development can be con-sidered as a signal of less close attention to themarket.

Furthermore, it should be observed that there is

Innovation Capabilities in Small Software Firms 345

TABLE IResources employed to evaluate the degree of market innovation capability and the degree of technological innovation capability

Resources linked to the DMIC Resources linked to the DTIC

C1: Entrepreneurial resources C1: Entrepreneurial resourcesC1.1: Number of persons forming the entrepreneurial C1.1: Number of persons forming the entrepreneurial

group groupC1.2: Entrepreneurs’ know-how C1.2: Entrepreneurs’ know-how

C1.2.2: Percentage of entrepreneurs with market C1.2.1: Percentage of entrepreneurs with knowledge technical knowledge

C1.2.3: Percentage of entrepreneurs with business C1.3: Involvement of entrepreneurs in software design and management experience and development

C1.3: Involvement of entrepreneurs in software design and development

C1.4: Involvement of entrepreneurs in marketing activities

C2: Resources linked to human resources C2: Resources linked to human resourcesC2.1: Total number of employees C2.1: Total number of employeesC2.2: Percentage of software developers C2.2: Percentage of software developersC2.5: Training C2.3: Percentage of internal software developers having

a graduate degreeC2.5.1: Marketing and management training

C2.6: Marketing aspects C2.4: Job rotationC2.6.1: Percentage of internal persons involved C2.5: Training

in marketing activitiesC2.6.2: Percentage of external persons involved C2.5.2: Technical training

in marketing activities

C3: Resources linked to external network C3: Resources linked to external networkC3.3: Intensity of commercial collaboration with C3.1: Use of non proprietary tool or external

other firms methodology of software developmentC3.2: Intensity of technical collaboration with other

firmsC3.4: Intensity of group relationships with R&D

centers, other software firms

C4: Economical indicators C4: economical indicatorsC4.1: Total profit deriving from firm’s software C4.2: Total profit due to non proprietary hardware andC4.3: Total profit coming from extra-regional market software

only a partial symmetry between the structures ofDTIC and DMIC. For example, while someresources used to evaluate the DTIC have theircounterpart in the DMIC structure, such as thepercentage of entrepreneurs with technical knowl-edge and the percentage of entrepreneurs withmarket knowledge, in other cases the market coun-terpart of a technical variable does not exist, as isthe case for the degree of involvement of entre-preneurs in software development. The reason forthis imperfect symmetry lies in the fact that thedevelopment of small software firms is usuallymainly technology-push rather than demand-pull.For example, since entrepreneurs usually have astrong technical background, their intense involve-ment in software development activity may implya lack of attention to the market. On the otherhand, an entrepreneur’s strong involvement inmarketing activities usually does not mean a lackof interest in technology, but rather a separationbetween the management and production activi-ties which are almost entirely delegated to spe-cialized personnel.

3. The proposed methodology

Within the theoretical assumptions described inthe previous sections, our methodology focuses ontwo main aspects: the description of a firm’s life(case study analysis) as succession of stablephases separated by critical events, and the eval-uation of resources managed by firms during eachstable phase by means of a formal model presentedin Section 4. The steps of the methodology canbe summarized as follows:

(a) Case study analysis;(b) Evaluation of resources;(c) Quantitative analysis of firms’ innovation

capabilities.

In the following a brief description of each stepis provided.

Case study analysis: in this step data and infor-mation concerning the firm’s history has to becollected through interviews with entrepreneurs.In particular, interviews should focus on how theinnovation capabilities of a firm have changed,and which kinds of resources have allowed thecompany to sustain innovation through time.

Through a case study approach, critical events inthe firm’s life that caused notable modifications ofresources, and stable phases, in which there areno significant changes, have to be identified.Examples of critical events are failure/initiation ofcollaborations with other firms (joint ventures,supply relationships), failure/initiation of rela-tionships with important customers, loss/acquisi-tion of critical professional resources because ofturn-over, or very rapid growth.

Resource evaluation: in this step the amount ofresources contained in Table I and available to thefirm in a given stable phase must be evaluated.Two critical points arise in this step: (i) entrepre-neurs are very often not able to remember theexact values exhibited by some resources in thepast, so they usually express only approximate andvague judgements; and (ii) some resources areintrinsically qualitative or hard to measure throughprecise indicators. In order to cope with uncer-tainty and vagueness deriving from ignorance orintrinsically qualitative assessments, the proposedmethodology allows entrepreneurs to express theirjudgements through verbal scales and to model, inthe next step, such qualitative assessments throughfuzzy set theory (Zadeh, 1973).

Quantitative analysis of the firm’s innovationcapabilities: In this step, data and informationcharacterizing critical resources are elaboratedthrough a mathematical model described in thenext section. The objective of this step is to deter-mine, in each stable phase, a quantitative evalua-tion of the innovation level achieved by a firm,starting from imprecise and qualitative judgementsexpressed by entrepreneurs and employees in theprevious step. By repeating the assessment foreach phase, it is possible to trace the market andtechnological innovation trends of a firm. Beforeapplying the proposed methodology to a set ofcase studies as shown in §5, in the next section weprovide a description of the mathematical modelfor the evaluation of innovation capabilities.

4. An evaluation model for the determinationof small firms’ innovation capabilities

In this section a mathematical model for evalu-ating the degree of technological innovation

346 Guido Capaldo et al.

and the degree of market innovation is presented.By using the proposed model, it is possible toevaluate the Degree of Technological InnovationCapability (DTIC) and the Degree of MarketInnovation Capability (DMIC) of a firm in eachstable phase, on the basis of the amount of theresources described in Section 2 and summarizedin Table I.

The model can be described with respect to thefollowing salient aspects:

(a) assessment of resources; (b) data elaboration.

4.1. Assessment of resources

A main issue in the construction of the model isto define a reliable procedure for assessing thevalues of the resources summarized in Table I. Itis evident that among them there are both quanti-tative and qualitative variables.

While, for example, it is easy to determinevalues such as the number of software developersor the amount of profit from extra-regionalmarkets, it is difficult to assess the intensity of thecollaboration with other firms or the degree ofinvolvement of the entrepreneurs in softwaredevelopment. Moreover, because of a lack ofinformation, a certain amount of imprecisioncould affect also crisp data. Chen and Hwang(1992), for example, identify four situations thatmay produce qualitative uncertainty: unquantifi-able information, incomplete information, non-obtainable information, and partial ignorance.

In order to cope with qualitative or impreciseinformation, the elicited variables are representedas linguistic or fuzzy variables, that is variablesassuming as their values linguistic terms such aslow, very low, more or less high representedthrough fuzzy sets (Zadeh, 1973).

Fuzzy representation helps to keep account ofthe ambiguity and vagueness embodied in quali-tative assessments. For example, if we considerthe variable intensity of technical collaborationwith other firms one could say that it is low oraverage or more or less high.

Because of their ambiguity, linguistic judge-ments can be represented in a more meaningfulway through fuzzy sets rather than through numer-ical scales. Also the imprecision contained instatements such as about five or six software devel-

opers or a high amount of my time was devotedto software design can be taken into account by afuzzy representation.

The rationale behind the fuzzy approach is thatthe transition between low involvement andaverage involvement is not sharply defined but is,rather, a matter of degree. This approach is muchmore valid than a binary point of view in manyreal-life situations, where it is often impossible todetermine what is the exact value marking theboundary between two categories by fixing anarbitrary threshold.

4.2. Data elaboration

In this model, the issue of determining values forthe DTIC and the DMIC is resolved through afuzzy multi-attribute decision-making approach.Generally speaking, this representation of theproblem can be synthesized as follows: given analternative A to be evaluated with reference to aset of n attributes or criteria {C1, C2, . . . , Cn},the overall evaluation g of A with respect to theCi’s can be determined in the following way:

g(A) = f [c1(A), c2(A), . . . , cn(A)] (1)

where ci(A) is the score obtained by A with respectto the criterion Ci and f is a monotonically non-decreasing aggregation function satisfying someintuitive criteria, usually represented in a formalway by means of a set of axioms.

In our representation A is the firm to be evalu-ated with respect to the DTIC or the DMIC, Ci(A)is the amount of one of the resources containedin Table I and managed by the firm at a certaininstant of time, ci(A) are fuzzy scores expressingthe values assumed by each resource obtainedfrom a fuzzy representation of the verbal termsused during the interviews.

Consequently, in order to calculate g(A)according to the (1), it is necessary to choose asuitable fuzzy multi-criteria aggregation operator.

Several operators have been proposed inliterature (Chen and Hwang, 1992). In our studywe have used a family of aggregation operatorscalled OWA, Ordered Weighted Average operators(Yager, 1988).

This choice can be justified on the basis of thespecific context of application. It is possible todemonstrate that an OWA operator can be associ-

Innovation Capabilities in Small Software Firms 347

ated to a fuzzy quantifier, i.e. a fuzzy set repre-senting a vague linguistic quantifier such as, forexample, many, most, almost all. In the DTIC andDMIC calculations we have used the quantifiermost. By aggregating single scores through thequantifier most, the evaluation of the DTIC or theDMIC is high if a firm is able to manage a highamount of most of the considered resources. In thisway we are able to bear in mind our ignorance ofthe way in which resources are employed andexploited by firms. In other words, the aggrega-tion mechanism is such that if a firm has most ofthe resources, it does not matter which in partic-ular, the firm’s innovation capabilities are high.

5. Case studies

In order to illustrate the application of theproposed methodology, and to show how it canbe used in the evaluation of the degree of marketinnovation capability and of the degree of tech-nological innovation capability, in this section theresults obtained from the analysis of three casestudies of small software firms are presented. Thisdescription is articulated in two steps:

(a) Analysis of firm’s history (case description):the material concerning the case studies wascollected though the examination of companydocumentation and through several interviewswith entrepreneurs and managers. The inter-views were aimed at identifying the firm’shistory through a narrative perspective, byemphasizing the main changes and turningpoints of the firm’s life. The entrepreneurfound it easy to describe her/his firm’s historythrough the concepts of stable phases andcritical events, as well as to identify resourcesthat were critical in each phase. The synthesisobtained by both interviews and companydocumentation was shared with the entrepre-neurs who were asked to validate the resultsthrough a careful analysis of the transcriptions.

(b) Evaluation of innovation capabilities: throughthe model presented in the previous section,the values of the degree of market innovationcapabilities and of the degree of technologicalinnovation capabilities are calculated for eachstable phase. Each phase is characterized interms of the amount and the kind of the

resources summarized in Table I and managedby the firm.

5.1. Case 1

5.1.1. Case descriptionThe firm was founded in 1980 by three entrepre-neurs with a good knowledge of the market andstrong management experience. The firm’s mainactivity was the development of software for ware-house and stock management. Its customers werelarge supermarkets. At the beginning, the firmseemed to place itself in an intermediate positionbetween market and technology, and the entre-preneurial group demonstrated a certain emphasison market expansion. The firm did not have anykind of external relationships with other technicalcenters; its workforce consisted of experienced butnot highly qualified software developers.

In the early stage of the firms’ life the entre-preneurs were essentially focusing on increasingtechnical innovation. The number of internal soft-ware developers increased considerably and thefirm also enjoyed the external technical support ofa software consultant. In the absence of innova-tive “home-grown” products, the firm was able tosurvive through an increase in profits from itscom-mercialization of software and hardware. Innova-tive effort was directed to the upgrading of existingproducts for the same market, rather than thedevelopment of new products or the introductionof the existing product to new markets. The entre-preneurs seemed to be aware that they had a highlyinnovative product and their primary objective wasto enlarge the firm’s presence on its currentmarket. As a result of these efforts the firm wasable to achieve a certain growth in profit due tosoftware development (about 30% of total profit).

The first critical event the firm had to face wasthe saturation of its market. This circumstanceinduced the entrepreneurs to shift their attentionto market expansion. In order to accomplish thisgoal, the company modified the amount and natureof its resources in order to develop its marketingcompetencies: a marketing expert was employedand, at the same time, there was a remarkable shiftin the firm’s emphasis away from technologicalinnovation (reduction of the number of softwaredevelopers; the entrepreneurs stopped collabo-rating with the software project team).

348 Guido Capaldo et al.

This reconfiguration of the resource mixmarked the beginning of a stable phase from 1983to 1990. In this period (’83–’90) the firm experi-enced remarkable growth (profits were 400%higher than 1983).

The critical event marking the beginning of thethird phase (’90–’93) is the loss of two experi-enced software developers. The strong orientationto the market was also confirmed in this phase,and was accompanied by a growing indifferenceto technical aspects (low involvement of entre-preneurs in software development, low intensityof technical collaboration with other firms).

As a consequence of an excessive focus on themarket aspects, the firm began to pay for its indif-ference to technological innovation: at the endof 1993 the firm’s profit diminished by 50%compared to 1990 because of the increasing obso-lescence of the firm’s software products.

5.1.2. Evaluation of innovation capabilities Figure 1 shows the innovation performance ofthe firm described in the previous paragraphthrough time. These figures have been obtainedby calculating the degree of technological andmarket innovation (respectively DTIC and DMIC)through several years, by means of the evaluationmodel presented in Section 4.

Case 1 represents an example of what we termmarket-oriented behavior. The firm described inthis case shows an very rapid initial growth of thetechnological innovation trend (LTI) due to thehigh innovative content of its early products,followed by a period of decline. In this case, theattention of the entrepreneurs to market aspectsallowed the firm to survive for a long period of

time (’83–’93) with good economic results, evenin the absence of substantial technological inno-vation efforts. Nevertheless, the strong marketorientation, despite the lack of a real technolog-ical innovation policy, was able to slow down butnot to prevent firm’s LTI decline.

In order to provide some details concerning theapplication of the methodology to the describedcase study we offer the following example inwhich we show how to calculate the DMIC in afixed instant of time.

ExampleIn order to calculate the value of DMIC and DTICfor a given firm, we need to assess the amountof resources listed in Table I available to thefirms in a given instant of time. For example,with reference to case study 1, we have the fol-lowing data as regards the DMIC in 1990 (seeTable II).

According to the proposed methodology, verbaljudgements are represented by means of fuzzy setscorresponding to a set of seven verbal terms {verylow, low, more or less low, average, more or lesshigh, high, very high}; verbal terms are repre-sented for each variable contained in Table Ithrough triangular membership functions definedon a specific universe U (e.g. the number of entre-preneurs is defined on the interval [0, 10]). Valuesare collected both from interviews and from theanalysis of company data. Judgements are thenconverted in crisp numbers (third column ofTable II) through the dual truth model (Zollo etal., 1999), which permits us to take account of theuncertainty of the judgement in determining the

Innovation Capabilities in Small Software Firms 349

Figure 1. Case-study 1 results.

crisp value, and aggregated through a fuzzyquantifier most according to the relation (1) inSection 4.

In the example we obtain an aggregated valueequal to 0.61; this result should be interpreted asthe degree of truth of the proposition “Most of theresources linked to the DMIC are available to thefirm in the instant t0”.

5.2. Case 2

5.2.1. Case descriptionThis firm was founded in 1979 by threeentrepreneurs with a high level of technologicalknowledge. Together with a group of external con-sultants, the entrepreneurs developed software forstructural engineering. Their strongly technicalculture together with their intensive involvementin software development demonstrated the firm’sstrong focus on technical aspects. At the sametime, the absence of entrepreneurs with manage-ment experience, and the lack of specificprofessional figures taking charge of marketingactivities, are clear signs that the firm did not paysufficient attention to the market in the firstphases. This tendency was also confirmed throughthe whole first period (’79–’83) by other factorssuch as the firm’s relationships with other tech-

nical centers, and the involvement in softwaredevelopment of ten external collaborators offeringtechnological consultation. In practice, everyoneworking for the firm (employees, consultants, andentrepreneurs) was involved in software designand development. This phase was characterized bya rapid growth in profit: in 1983 the firm doubledits profit compared to the previous period.

The critical event marking the beginning of thesecond phase from 1983 to 1990 was the recruit-ment of several new software developers and aconsequent significant increase of the company’snumber of employees. Notwithstanding this event,in this phase, there was an evident decreasein innovative effort since the company seemedcommitted to a more intense commercial exploita-tion of its products and, at the same time, theentrepreneurs seemed to pay more attention tomarket development than in the previous phase.This phase was also characterized by a significantdiversification of software products throughthe implementation of software packages foraccounting and management. The organizationalstructure is characterized by small project groupsworking in the firm’s two main businesses: engi-neering support software and managerial software.The decrease in priority given to technologicalinnovation was not accompanied by an effective

350 Guido Capaldo et al.

TABLE IIApplication of the proposed methodology: An example

Resources linked to the Degree of Market Innovation (1990) Evaluation Fuzzy value

C1: Entrepreneurials resourcesC1.1: Number of persons forming the entrepreneurial group High 0.725C1.2.2: Percentage of entrepreneurs with market knowledge More or less high 0.615C1.2.3: Percentage of entrepreneurs with business and management experience High 0.725C1.3: Involvement of entrepreneurs in software design and development Very Low 0.165C1.4: Involvement of entrepreneurs in marketing activities Very high 0.835

C2: Resources linked to human resourcesC2.1: Total number of employees Average 0.5C2.2: Percentage of software developers Low 0.275C 2.5.1: Marketing and management training Average 0.5C 2.6.1: Involvement of internal persons in marketing activities High 0.725C 2.6.2: Involvement of external persons in marketing activities Low 0.275

C3: Resources linked to external networkC 3.3: Intensity of commercial collaboration with other firms High 0.725

C4: Economical indicatorsC 4.1: Total profit deriving from firm’s software Average 0.5C 4.3: Total profit coming from extra-regional market High 0.725

policy of market expansion, though there was acertain increase in this direction. In other words,the company management perceived the difficultyof enlarging the market but tried to solve thisproblem through a re-engineering of internal orga-nization and procedures, instead of initiating aneffective policy of market expansion.

A change in the entrepreneurial group becauseof the departure of one of the founders was thecritical event marking the beginning of the thirdphase (’90–’93) in which the same tendency of theprevious phase seemed to be confirmed. The maincustomers were banks (40%) and engineeringsocieties (40%); a certain amount of profit derivedfrom software development for small firms (20%).In this phase, again, the company managementtried to modify the organizational structure by theintroduction of a divisional structure made up oftwo business units: bank services and technicalservices. Notwithstanding the introduction of inno-vative methodologies in software development andin project management, the overall innovative per-formance remained quite low. In fact, in about tenyears the firm only introduced two new products,followed by their upgraded releases onto themarket. Consequently, in this period the firm’sprofit remained unaltered. The redefinition andrationalization of the organizational structure aswell as the introduction of innovative methodolo-gies in software development and project man-agement were not able to ensure an increase in theinnovative performance as they were not accom-panied by real and effective product innovationand market expansion.

5.2.2. Evaluation of innovation capabilitiesThe evaluation of DMIC and DTIC for this firmis depicted in Figure 2.

This case represents an example of what weterm technology oriented behavior; the firm showsa strong initial orientation toward technologyfollowed by a rapid decline of its innovativecapabilities. The substantial difference with theprevious case study lies in the fact that this firmwas not able to compensate the low level of tech-nological innovation by developing an effectivepolicy for enlarging its market. A small firm fol-lowing this pattern may perish or may survive ina very limited market niche (local or linked to avery specialized product), as happened in thiscase (extra-regional profit showed a drasticdecrease). This case also shows that innovationin organizational structures and in design method-ologies not followed by an adequate innovationtransfer to new products is not sufficient toproduce an increase of the firm’s innovationcapabilities.

5.3. Case 3

5.3.1. Case descriptionThis firm was founded in 1978 by four entrepre-neurs. In this case too, the entrepreneurial groupwas characterized by a strong technical culture. Inthe first phase (’78–’82) the firm seemed to attachgreat importance to technological innovation. Theentrepreneurs were totally involved in the activityof software design and implementation. The wholeorganization was composed of the entrepreneurialgroup and one external software collaborator. Thefirm’s market was limited to a local dimension.

Innovation Capabilities in Small Software Firms 351

Figure 2. Case-study 2 results.

About 35% of the firm’s profit came from otheractivities that were not linked to software pro-duction.

The passage to the second phase (’82–’83) wasmarked by the perception by the company’s man-agement of an evident market saturation immedi-ately followed by a certain increase in the firm’sattention toward marketing aspects. There was asubstantial growth of the number of employeesfrom 4 to 8. Employees were exclusively devotedto software development, while entrepreneurs par-tially took charge of marketing aspects. The entre-preneurs still maintained the direction of theproject team. The growth of the firm’s profit wasrather consistent in this phase.

The third phase (’83–’88) was characterized bythe acquisition of an important technical partner(a large software firm) and some new importantcustomers among the largest public and privateItalian companies. Those events provoked arenewed interest in technological innovation,thanks to the acquisition of important know-howfrom the new partner, and a remarkable generalgrowth of the firm. The number of employeesdoubled compared to the past and, as in previousphases, most of them were highly qualifiedsoftware developers. At the same time, the entre-preneurs were no longer involved in softwaredevelopment and their role became essentiallymanagerial, though still anchored in a technicalgrounding. An intermediate organization level wasformed and the coordination of the project teamwas delegated to a project manager who did notbelong to the entrepreneurial group. A certaindegree of flexibility was achieved by an intensivejob rotation policy.

In the fourth phase (’88–’93) the firm againshowed a closer attention to the market and alower innovative effort than in the previous phase.The critical event was the starting of a proactivemarketing policy by means of the creation of anautonomous marketing unit. The firm’s objectivewas twofold: to enlarge profits in new extra-regional markets considerably, and to diversify theproduct portfolio by moving towards providingprofessional services such as training programs toother companies. This phase was characterized bya striking growth of the firm both in profit and inworkforce. The new employees continued to besoftware developers, though less qualified than

the older ones. From the organizational point ofview there were substantial innovations. The firmpresented an articulated organizational structuredesigned to enhance and develop technologicaltransfer through a high degree of flexibility andan intensification of the relationship betweenmarketing and R&D. In order to accomplish thisresult, the organizational structure was character-ized by a matrix model, which related the firm’sbusiness units to competency-based professionalareas. Also, in this phase, intensive research intoinnovative software development and analysismethodologies was carried out by the firm, whichwas involved in a national research program intosoftware engineering. It is worth noting that inall phases, intense internal training activity wastaking place.

5.3.2. Innovation capabilities evaluationThe results concerning the evaluation of theDMIC and the DTIC for this firm are shown inFigure 3.

This case is an example of what we term oscil-lating behavior: the firm is able to alternate phasesin which there is a strong attention to technolog-ical innovation with phases in which marketaspects are given greater prominence.

6. Discussion of results

In this section we present a comparison among thethree case studies in order to better highlight someresults emerging from the field study, concerningthe analysis of innovation capabilities.

In Case 2 (technology-oriented case), the initialgrowth ends because of the firm’s inability toexploit the market potentialities of its products;in Case 1 (the market-oriented behavior) the firmis able to expand the market but not to maintaina satisfactory level of technological innovation. Inboth cases, firms are confined to a very limitedmarket niche. The main difference in the third case(oscillating behavior) compared to the first twolies in the fact that, in the market phases, the firmdid not neglect a technological focus, and was ableto achieve a certain balance between market andtechnology.

All the considered firms show some commonfeatures. They were founded by technical entre-preneurs and were born around a product idea

352 Guido Capaldo et al.

developed for a specific market, often limited toa regional market. The acquisition and the updatingof initial know-how took place in several ways:through relationships with large firms or by meansof close relationships with technical groups orresearch centers, which allowed the firm to exper-iment with high levels of technological innovationand specialize in a well-defined market segment.The field analysis showed that all the firms,because of the high degree of technical expertiseof the founders and their network of relationshipswith other technical centers, reached a good levelof technological innovation (LTI) in the first yearsof their life. This tendency is shown in all thepresented cases in which LTI is always more orless high in the first phase or at the firms’ birth.

Nevertheless, the strong focus of a group on thetechnological side usually implies neglectingmarket vision and development (customer fidelity,marketing and commercialization, services andassistance, connections with other firms). For thisreason all the considered firms experience a moreor less dramatic market crisis, here essentially seenas an inability to enlarge the market beyond thelocal area and a difficulty in maintaining growth.This crisis is particularly evident in the first twocases (see the LMI graphs).

The market crisis causes a reorganization of thefirms’ activities, involving both the original groupof entrepreneurs and the internal professionalskills. In general, one or more entrepreneurs movefrom product development activities towardsmarket development or managerial roles.

The reorganization of the firm following themarket crisis signals the beginning of the declinein a firm’s innovative ability. Firms reduce their

engagement in product development and progres-sively concentrate on the supply of IT services orlimit themselves to updating the initial product, asin Cases 1 and 2, in which it is possible to observea remarkable fall of the LTI while LMI increasesor remains stable. In Case 3, the firm managed tointerrupt the decline with a reorganization of activ-ities to take advantage of opportunities arisingfrom a new relationship with the market. This firmshows continuous progress in innovative activities(the LTI increases again after the initial fall).

7. Conclusion

In this paper we have presented a methodology toevaluate the innovation capabilities of smallsoftware firms. The main hypothesis on which ourapproach is based, i.e. the presence of a relation-ship between innovation capabilities and certainkinds of resources managed by firms, has a doublesource: the resource-based theory and the litera-ture on small software firms, based on empiricalobservations concerning the variety of resourcesmanaged by firms and the transience of their inno-vation capabilities.

Putting together theoretical models and empir-ical observation has turned out to be a critical taskfrom the methodological point of view because ofthe difficulty of providing a clear identificationof firms’ resources. The strong specificity ofresources related to innovation capabilities haveconvinced us of the need to restrict the analysisto a small group of firms operating in the samesector to be analyzed as case studies. In this waywe have been able to concentrate our attention onthe methodological aspects.

Innovation Capabilities in Small Software Firms 353

Figure 3. Case-study 3 results.

Starting from the methodology presented in thispaper, it should be possible to develop tools forthe support of innovation management for smallsoftware firms, e.g. to identify which resourcesshould be acquired, or to solve the dilemma ofhow many resources to use on the technologicalfront and how many for the market. In order toaccomplish this goal however, empirical analysison a large sample of firms should be carried out,to test the efficacy of the methodology and totransform it from a descriptive to a prescriptivemethodology.

As regards the possibility of generalizing theresults of the field analysis, the three kinds ofbehaviors (technology-oriented, market-orientedand oscillating) have also been detected inother case studies that are not presented here.Notwithstanding this, in order to understand ifthese three behaviors correspond to general trends,large sample and cross-sector studies should beperformed.

Note

* Although the paper is the fruit of the collaboration of allthe authors, in this version, Section 1 is by M. Raffa, Section2 is by G. Zollo, Section 3, 5.1.1, 5.2.1, 5.3.1 are by G.Capaldo and Section 4, 5.1.2, 5.2.2, 5.3.2 are by L. Iandoliand remaining parts are the fruit of collaborative work.

References

Amit R. and P. J. H. Schoemaker, 1993, ‘Strategic Assets andOrganizational Rent’, Strategic Management Journal 14,33–46.

Barney J. B., 1991, ‘Firm Resources and SustainedCompetitive Advantage’, Journal of Management 17,99–120.

Barocci, T. A., K. R. Wever and R. A. Lahey, 1983, ‘HumanResource Planning for Information Systems Personnel:Skills Mixes and Technological Trends’, working paper1478–1483, Sloan School of Management, MassachusettsInstitute of Technology.

Bellini E., G. Capaldo, M. Raffa and G. Zollo,1997, ‘University as Resources for Entrepreneurship’,Proceedings of UIC/AMA Symposium on Marketing andEntrepreneurship, Chicago.

Churchill N. V. and V. L. Lewis, 1983, ‘The Five Stages ofSmall Business Growth’, Harvard Business Review 3,30–50

Conner K. R., 1991, ‘A Historical Comparison of Resource-based Theory and five Schools of Thought withinIndustrial Organization Economics: Do we have a newTheory of the Firms?’, Journal of Management 17,121–154.

Chen S. J. and C. Hwang, 1991, Fuzzy Multiple AttributeDecision Making, Berlin: Springer Verlag.

De Chiara A., 1996, ‘L’applicazione della teoria delleresource-based strategies nella piccola impresa: risorseinterne ed attitudine al cambiamento aziendale’, PiccolaImpresa/Small Business 2, 52–70.

Garden A. M., 1992, ‘Potential Reasons for SoftwareEmployees in Small Companies to Leave Their PresentCompany’, IEEE Trans. Engineering Management 39(3),246–253.

Grant R. M., 1991, ‘The Resource-Based Theory ofCompetitive Advantage: Implications for StrategyFormulation’, California Management Review, Spring,114–135.

Greiner L. E., 1972, ‘Evolution and Revolution asOrganizational Growth’, Harvard Business Review 4,98–114.

Kelley M. R. and H. Brooks, 1991, ‘External LearningOpportunities and the Diffusion of Process Innovations toSmall Firm’, Technological Forecasting and SocialChange 39, 27–41.

Lippit G. and W. Schmidt, 1967, ‘Crises in DevelopingOrganizations’, Harvard Business Review 5.

Miller D. and P. H. Friesen, 1984, ‘A Longitudinal Study ofthe Corporate Life Cycle’, Management Science 30,1161–1183.

Quinn J. B., 1979, ‘Technological Innovation, Entrepreneur-ship and Strategy’, Sloan Management Review 20(3),102–128.

Raffa, M. and G. Zollo, 1988, Software: Tecnologia eMercato, Bologna: Il Mulino.

Raffa M. and G. Zollo, 1992, ‘Il Rapporto tra InnovazioneTecnologica e Organizzazione nelle Piccole ImpreseInnovative’, Piccola Impresa/Small Business 2, 49–72.

Raffa M., G. Zollo and R. Caponi, 1996, ‘The DevelopmentProcess of Small Firms’, Entrepreneurship & RegionalDevelopment 8, 359–371.

Raffa M. and G. Zollo, 1998, Economia del Software, Napoli:ESI.

Reid G. C. and L. R. Jacobsen Jr., 1988, The SmallEntrepreneurial Firm, Aberdeen (UK): University Press.

Rumelt R. P. 1987, ‘Theory, Strategy, and Entrepreneurship’,in D. J. Teece (ed.), The Competitive Challenge,Cambridge (MA): Ballinger, pp. 137–158.

Steinmetz L., 1969, ‘Critical Stages of Small Business’,Business Horizons 12(1).

Wiklund J., 1996, ‘Theoretical Perspectives in Research onFirm Growth’, RENT X, Brussels, November 20–23.

Yager, R. R., 1988, ‘On Ordered Weighted AveragingAggregation Operators in Multi-criteria Decision-making’,IEEE Transactions on Systems, Man and Cybernetics18(1), 183–191.

Zadeh, L., 1973, ‘Outline of a New Approach to the Analysisof Complex Systems and Decision Processes’, IEEETransactions on Systems, Man and Cybernetics 3(1),35–48.

Zollo, G., Iandoli, L., Cannavacciuolo, A., 1999, ‘The per-formance Requirements Analysis with Fuzzy Logic’, FuzzyEconomic Review 4(1), 35–73.

354 Guido Capaldo et al.