Strategic Management JournalStrai. Mf-mr. J.. 25: 85-99 (2(M)4)

Published online in Wiloy InlerSL'ioticc (www.intersL-ience.wiiey .cam). DOI: l().l{l()2/smi.365

STRATEGIC POSITIONING, HUMAN CAPITAL, ANDPERFORMANCE IN SERVICE ORGANIZATIONS:A CUSTOMER INTERACTION APPROACHBRUCE C. SKAGGS'* and MARK YOUNOr' Gatton College of Business and Economics, University of Kentucky, Lexington,Kentucky, U.S.A.^ Department of Management and Business, Skidmore College, Saratoga Springs,New York, U.S.A.

This study employs a customer interaction approach to examine how service organizations'strategic positioning relates to their human capital, and how the interaction between strategicpo.sitioning and human capital impacts organizational performance. Results from 234 serviceorganizations in 96 different indu.strics indicate very strong relationship.s between strategicpositioning choices ami human capital. We a I.so find that certain comhiiiations of strategicpositioning and human capital result in superior performance. Ct)pyriyht © 2003 John Wiley &.Sons. Ltd.

There has been a great deal of discussion in theliterature concerning the nature of service firms(e.g.. Brush and Anz. 1999; Lovelock and Yip,1996: Mills. 1986; Nayyar. 1993; Normann, 19K4).Some scholars suggest that these organizations areunique {i.e., different than manufacturers) and, assuch, require the developtnent of new models inorder to further our understanding of them (e.g..Mills cl at., 1983). Others believe that any differ-ences that exist do so in degree rather than in kind.und therefore existing organizational models arerobust enough to incorporate any nuances foundin Ihe service sector of the economy {e.g.. Bharad-waj. Varadarajan. and Fahy, 1993). Regardless ofthis debate, there does appear to be general agree-ment that differences do exist between service andmanufacturing in a few specific areas. One of themost often cited of these is customer interactionwith the production process (Kotler, 1983; Nor-mann. 1984).

Key words: strategic positioning; services; human capital*Ci)rresp(indence lo: Bruce C. Skaggs. Gatton College of Busi-ness and tiuonomics. University of Kentucky. Lexington. KY40506. U.S.A. E-mail: bskag2^>uky.edu

Accordingly, scholars have begun to focus onhow the interaction between the customer and thefirm influences eletnents of the production pro-cess. For example, in an examination of emer-gency room units. Argote (1982) found that thegreater the variety of patient conditions treated,the more nonprogrammed the decisions are dur-ing service production. This finding is consistentwith Jones" {1987) examination of small, mid-western service firms where he found a positiverelationship to exist between the level of uncer-tainty surrounding the customer-firm interactionand the level of service production complexity.Furthermore, each of these authors strongly indi-cates that strategy acts as the driving force behindthe level of customer-induced uncertainty. That is.the service firm's strategic positioning dictates thetype of information tequired from customers dur-ing the interaction and, hence, the correspondinglevel of uncertainty that confronts the productiiinprocess (Argote. 1982; Jones. 1987). Thus, thesestudies support the idea that sttategic positioningchoices influence the level of uncettainty that cus-tomers bring to the organization's production pro-cess. In response, service firms alter elements of

Copyright © 2003 John Wiley & .Sons. Ltd. Received 28 September 2001Final revision received 30 June 2003

86 B. C. Skaggs and M. Youndt

their production processes in otder to effectivelycope with the level of uncertainty generated bythe customer interaction (Jones. 1987; Mills andMoberg, 1982; Shostack. 1987;Tansik, 1990).

Of the elements that comprise service produc-tion, humun capital, or the skills, knov^ledge. andexpertise of employees (Becker, 1964. Schultz.1971), is believed to be one of the most impor-tant (e.g.. Mills. 1986: Quinn, 1992). Unlike thatin manufacturing firms, customers of service firmstypically interact vt'ith the production process (Nor-mann. 1984). And since production in servicesis typically dominated by labor, customers havea high degtee of interaction with the productionemployees of the firm (Mills. 1986). Moreover,any uncertainty that customers interject into theprocess is being confronted in large part by theseproduction employees (Mills and Morris. 1986).Thus, if service firms alter elements of the pto-duction process to handle varying levels of strate-gically induced customer uncertainty as previousstudies have suggested (e.g.. Jones. 1987). thenthis has important implications for human capitalinvestment decisions.

Unfortunately, very little attention has heendirected at exploring the interaction of strategicpositioning and human capital in service organi-zations. Furthermore, of those studies that haveexamined this relationship (e.g.. Boxall and Steen-eveld. 1999; Delery and Doty. 1996; Hitt el al.,2001). none have considered the impact of the cus-tomer interaction. While these studies are impor-tant first steps, the general dearth of research inthis area is a bit surprising given the dependence ofservice organizations on human capital. Moreover,since employees of service organizations tend to belinked to customers during service delivery, theyplay a key part in carrying out the strategic initia-tives of these firms. Thus, we believe that by takinginto account the presence of the customer duringservice production, we can yield valuable insightsfor researchers and practitioners concerning thelinkages among strategic positioning, human cap-ital, and organizational performance in serviceorganizations. As a result, managers of these firmsmay be better able to implement strategic deci-sions as well as make more effective investmentsin human resources. Likewise, researchers wouldbe granted a better understanding of the dynamicsdriving human capital decisions in service firms, asector of the U.S. economy which now represents

over 75 percent of both employment and GPD(United States Bureau of Labor Statistics, 1997).

In the following sections we begin by arguingthat different strategic positioning choices impactthe customer interaction with the firm. In particu-lar, we suggest that strategic positioning choices byservice firms will influence the potential variancein customer demands (i.e.. the range of customerneeds) that clients impose on a firm's productionprocess, which in turn requires varying levels ofhuman capital in order to deal with the correspond-ing uncertainty. Using this notion of variance incustomer demands as the organizing principle, wethen examine the specific relationships betweenstrategic positioning and human capital in ser-vice firm production. Next, we explore whetherthe hypothesized relationships between strategicpositioning and human capital yield superior orga-nizational performance. Lastly, we discuss howmanagers and researchers can use our findings intheir organizations and future research endeavors.

THEORY AND HYPOTHESES

Customer interaction, strategic positioning,and human capital

As customers of a service firm interact with itsproduction process, the variance in their demandscreates uncertainty for the organization (Argote.1982; Mills and Moberg, 1982; Tansik. 1990). Inresponse, service organizations may alter aspectsof production in order to process the requiredlevels of uncertainty. As stated above, a majorelement of service production is human capital.When customers introduce a high degree of vari-ability into the service production process, serviceorganizations may be able to address this vari-ability (i.e.. successfully satisfy customer needs)wben their employees are proficient at diagnosingproblems, thinking creatively, developing novelsolutions, and so on; that is. when they possesshigh levels of skill, knowledge, and expertise (i.e..human capital). This is consistent with human cap-ital theory, which contends that employees withhigh skill levels can better cope with uncertainty inthe task environment (e.g., Becker, 1964; Schultz,1971; Snell and Dean. 1992). In the present case,this uncertainty is the result of variability in cus-tomer demands (Jones. 1987).

Though variability in customer demands mayinfluence human capital requirements, it has been

Copyright -ij 2IH):i John Wilev A Sons. Lid. Mam. ./.. 25: H5-W (2()(I4|

Strategic Po.siti(mini>, Htunau Capital, and Performance 87

suggested that this variability is the result of strate-gic positioning decisions on the part of servicetirms (Jones, 1987; Tansik, 1990). According tothe positioning view of strategy. Hrms undertakestrategic positions in order to differentiate them-selves from existing and potential competitorsalong dimensions that are of importance to cus-tomers (e.g.. Porter, 1980. 1996). Since much ofwhat customers purchase frotn a service firm is aprocess (Normann. 1984). and since they are inter-acting with that process, differences in the produc-tion process itself allow for differentiation amongcompeting tirms (Mills, 1986; Shostack. 1987).Three key differentiating mechanisms involvingthe production process that have been discussed inthe services literature are cit.stonier co-produclion.customer contact, and service customization (e.g.,Hcskett. 1986; Mills. 1986). All service Hrmsundertake each of these to varying degrees (Nor-tuann. 1984). Furthermore, since these mecha-nisms are difficult for service fums to change inthe short term (Mills. 19H6), they represetit strate-gic elements to these organizations.

Customer co-production

As customers actively engage in the productionol the service, they are expending efft»n. Ser-vice firms can alter this aspect of service cre-ation, known as customer co-production, to differ-entiate themselves from competitors by offeringtheir services in a tiianner that requires greatert)r lesser amounts of effort on the part of cus-tomers (an option which very few manufacturingfirms possess) (cf. Maister and Lovelock. 1982;Mills. 1986; Upah. 1980). For example, some ser-vice firms require custt)mers to wait for service(first come, first served), to receive service onlyat a specific time (3:00 check-in), or to performvarious tasks (e.g.. bus tables, carry your ownluggage). Tbese are instances of high customerco-production. Other service organizations tnaychoose to position their service in a matiner thatreduces customer effort (e.g., free pick-up service.use of a loaner car. meals provided on a Might).

As a firm utilizes higher levels of co-productionit tends to reduce the range of customer needsthat it must address (e.g.. Mills and Morris. 1986).For example, some tax preparation firtns mayhave customers sort receipts into predefined cate-gories, thereby diminishing the potential breadth ofdemands customers may impose on the firm. The

use of these predefined categories creates a routineinput for the organization, reducing the amount ofuncertainty that is encountered in the task environ-ment during service production (e.g.. Scott. 1992;Thompson, 1967). This reduction in uncertaintyaffords the firm a relatively predictable task envi-ronment, which allows it to codify and simplifymuch of its productittn process. As a tcsull. weargue that the level of human capital required ofa service firm's production employees would bereduced. Accordingly, we anticipate:

Hypothesis I: Customer co-production will henegatively related to human capital.

Customer contaci

Whereas customer co-production concerns thelevel of effort customers are expending duringservice production, customer contact refers to thedegree of interaction they have with a firm'sproduction ptxKess (Mills, 1986). As firms increasethis degree ot interaction, they can decreasetesponse time in addressing customer needs.Howevet\ greater levels of interaction heightenthe information Hows between the customer andthe organization, providing customers with amechanism for injecting substantial amounts ofunceitainty into the service production ptocess(Mills, 1986; Mills and Morris. 1986; Tansik.1990). Fuiihertnore. as the level of interactionincieases, more areas of the production processcan become exposed to the customer, therebyincreasing the number of points in which thecustotner may inject variability. Together, wesuggest these will increase the need for effectiveinformation processing during service production.More specifically, as firms increase the level ofcustomer contact, we suggest their productionemployees will need a more enhanced skillset for dealing with customers, understandingidiosyncratic situations, making quick decisions.and so on. Conversely, if the service productionprocess is designed to minimize customer contact,less human capital will most likely be required. Asstich. we expect;

Hypothesis 2: Customer contact will he posi-tively related to human capital.

Copyrighl © 2l)f)_l John Wiley & Sons. UU. Strut. J- 25: 85-99 f21)04j

88 B. C. Skaggs and M. Youndt

Sen'ice customization

Serviee tirms also differentiate themselves in thecompetitive marketplace by the extent to whichthey customize the service offering tor individ-ual customers (Heskett. 1986; Normann, 1984).As a firm increases the possibility of meetingeach customer's specific needs, it faces greaterlevels of potential demand variability, which inturn increases the level of uncertainty that isencountered during service production (Tansik,1990). Accoidingly. we believe the decisions afirm makes concerning customization will in turnaffect its human capital considerations. High lev-els of customization requite production employeesto ascertain varying customers* needs, decide onthe particular procedures required to fulfill thoseneeds, and choose the most appropriate produc-tion path through the organization (Sht)stack. 1987;Tansik. 1990; Wathen.'l995). On the other hand.low levels of customization allow for service pro-duction to be routinized, thereby reducing the needfor Judgment on the part of production employees.In short, as customization incieases we expect ser-vice firms to cope with this strategically induceduncertainty in the production process through theuse of greater levels of human capital. Therefore,we anticipate:

Hypothesis 3: Service customization will he pos-itively related lo hutuan capilal.

Strategic positioning, human capita), andperformance

Service organizations are open social systems thatinterface with internal and external sources ofuncertainty (Argote, 1982). For the productionprocesses of service firms, a large part of thisuncertainty lies in their interactions with customers(Argote. 1982; Jones, 1987; Tansik. 1990). Wehave hypothesized that ser\ice firms may invest inhuinan capital so that their employees are capableof dealing with the potential variance in customerdemands generated by their strategic positioning.However, investing in human capital can be acostly endeavor (e.g.. Tushman and Nadier. 1978).The greater the variance in customer demands,the more sophisticated the skills needed to trans-form a wider variety of inputs into the serviceoffering. This sophistication gives rise to increasedcosts in terins of factors such as higher levels oftraining, higher levels of education, and the like.

Given the costs, it is suggested that efficient firmswill make just enough of an investment in humancapital to handle a given level of customer variabil-ity (cf. Williamson. 1979). This is consistent withAigote's (1982) finding that hospital emergencyrooms that achieved a fit between the level of coor-dination and the variance in customer demand gavemore effective treatment. Therefore, we suggestthat service firms that possess a match betweentheir strategic positioning and human capital willhave greater performance; that is. human capitalwill moderate the relationship between strategicpositioning and finn pertbrmance. Thus, based onthe expected associations between strategic posi-tioning and human capital detailed in HypothesesI, 2, and 3, we suggest the following moderationrelationships;

Hypothesis 4: Htintan capital will negativelymoderate the relationship between customer co-production and firm performance.Hypothesis 5: Huinan capital will positivelymoderate tiie relationship hetween cu.stomercontact and firm performance.Hypothesis 6: Human capital will po.sitivelymoderate the relationship hetween service cu.s-tomization and firm performance.

METHODS

Sample

In constructing the study sample, we consideredonly those service firms that were publicly heldin order to obtain objective performance data.We then used a size criterion of greater than$10 million in sales and larger than 50 employ-ees in order to increase the likelihood the orga-nizations possessed somewhat fbtmaiized strategicand HR activities. Finally, we limited our sam-ple to only those firms that teceived a majorityof their sales from a single industry to reducethe possibility of multiple strategic positioning andhuman capital investment choices occurring withina lirm. This scteening process yielded 1904 serviceorganizations. Questionnaires were mailed in late1997 and early 1998 to the highest-ranking exec-utive in each firm (usually the CEO. but in a fewcases the President or Chief Operating Officer).Though some have voiced concern over the useof a single respondent (e.g.. Simon and Burstein.

Copyright © 2003 Jotin Wilijy & Sons. Lid. Slruh t. J.. 25: 85-99 (20041

Strategic Positioning. Hunmn Capilal. and Petfonnance 89

I9K5). our decision to use a single respondent wasbased on two characteristics of the present study.First, all firms predominately cotiipete in a singleindustry. Thus, top mangers should have a greaterunderstanding of their firm's competitors, indu.strydynamics, and their own strategic positioning thanwould managers of diversified oigani/ations. Sec-ond, because service firms possess relatively highconcenttations of labor (i.e., labor is a primary firmresource) (Mills. 1986). it is likely thai top man-agers of these types of organizations will have anin-depth knowledge of their human capital. Thus,we believe these chatacteristics help to minimizethe concerns of using a single respondent.

A total of 234 organizations representing 96four-digit SIC codes tetutned questionnaites (seeTable I for a listing of the satnple by industrygroup). An analysis of respondent-nonrespondentorganizational differences based ou the numberof employees and performance showed the twogroups were not significantly different on eitherof these dimensions.

Measures

A pilot study involving 23 service timis fromfour industries was conducted to ascetiain thevalidity and reliability of tbe survey instrument.Sitiiilar to the main study, all firms had greaterthan $10 million in annual sales and 5i) or moreemployees. Alterations and clarifications weremade to the questionnaire items based on responsesand input from the pilot study. Questions forbotb the pilot and main studies were based on

seven-point Liken scales, and the pilot studyorganizations were not included in the sample ofthe main study.

Strategic positioning

These vatiables consisted of cu.stomer co-production, custotner contact, and .service cus-totnization and were nieasuted telative to cotii-petitors. Cusiotner co-production is the degree towhich customers participate in the design anddelivery of the service offering (Maister andLovelock. 1982: Mills and Morris. 1986) and wasmeasured using a Hve-itcm scale (alpha = 0.77).Cu.stomer cotitaci tefers to the degree to whichcustomers and employees interact with one anotherduring the service production process and wasmeasuied with a two-item scale (alpha = 0.95).Lastly, service customization refers to a lirm'sability to alter service production in order tomeet specific customer needs and was measuredusing a five-item scale (alpha = 0.89). See theAppendix for a list of the specific strategic posi-tioning variable itetiis.

Organizatiottal performance

Performance was measuted using two variables:teturn on equity (ROE) and return on investment(ROD. A 2-year average of both ROE and ROIwas used in order to reduce the potential forantimalous timing effects on performance. Eor ouranalysis, we combined ROE and ROI into t)ne

Tahic I. Correlalion.s, means, and siandarLi Jcviatidns

Variables

1. Si/.c2. Indtistry pcrforniancc.V Industry beta4. Information asymmetry5. Customer co-production6. Customer contaci7. Service eustomi/alion8. Human eapilal9. Perlormanee

Mean s.d.

3.13 0.78-0.33

0.95 (3.733.403.873.993.900.01 (

.70).3!2

.63

.33

.97

.58

.14).8S

1

0.14

-0.16-0.09

-0.10-0.02-0.08-0.06

0.02

0.16-0.01

0.02-0.10

0.02-0.03-0.01

3

-0.10-0.01

0.010.060.130.04

4

0.04-0.02-0.15-0.20

0.01

5

-0.040.02

-0.28-0.12

6

0.450.29O.(W

7

0.32-0.06

8

0.27

Correlations greater than 0.13 are signiCiCimt at /; < 0.05: correlations greater than 0.17 are significant at p < 0.01' A s recommended by Aiken and West (I99IK the strategic positioning and HRM variables were centered ( m e a n — 0 ) for ourmiideralod hierarchical regression interaction tests.Sample by indiislry group: Transporlation 118). Communications (3). Utilities {\5i. Wholesale Trade (25), Retail Trade 162), Insuranceand Real Estate (28). Personal iuid Business Services (38). Repair (2). Motion Picture and Amusement (10), Health, Educalion.Accounting and Related Services (33).

Cnp>rjghi Julm Wjk'v & Si.ins. Lid. ./.. 25: 85-W (2(X14)

90 B. C. Skaggs and M. Youndt

overall performance variable (alpha = 0.72). ROEand ROI data were obtained from Disclosure.

Human capilal

Organizations can increase their human capitallevels through selection and/or training (i.e., buyand/or make). Selection refers to an organization'sefforts to increase its human capita! by hiringemployees with high levels of education and exper-tise from the extemal labor market. Training, onthe other hand, refers to an organization's effort loincrease its human capital through internal devel-opmental activities of its current employees (Snelland Dean. 1992). While human capital theory canbe used to argue that selection and training aresubstitutes for one another and therefore redun-dant when used together, many studies have foundthem to be complementary (MacDuffie. 1995: Pfef-fer. 1994: Youndt ei ai. 1996). It is logical toconceive, for example, that organizations can useselection to increase their generic human capital,while focusing on training to develop firm specifichuman capital.

Following the complementary approach com-monly employed by strategic human tesource man-agement researchers studying high-performancework systems (e.g.. Arthur. 1994: Huselid. 1995:MacDuffie. 1995: Youndt etal., 1996), we useda composite selection and training scale consist-ing of five items to assess the level of humancapital possessed by the organization's employ-ees directly involved in ihe service production anddelivery process (alpha = 0.85). More specifically.the items focused on whether or not firms selectproduction employees with high levels of prioreducation, training, and experience, and how muchtime and money they spend on internal trainingactivities. Similar to the strategic positioning mea-sures, these items were assessed relative to com-petitors. See the Appendix for the specific humancapital items.

Control variahles

Previous studies have indicated that firm size andindustry environment can influence firm perfor-mance (e.g.. Keats and Hitt, 1988). To control forthe effects of siz.e, we gathered data on the num-ber of individuals employed by each firm. Numberof employees was used rather than total revenue,as researchers believe it to be a more accurate

proxy of size in the service sector (cf. Normann.1984). We then calculated the log of this numberand entered it in the fust step in our regressionanalyses.

ln order to control for industry effects, we firstcollected data using Research Insight on aver-age ROE and ROI for each of the 96 four-digitSIC codes represented in the study. We com-bined these into an overall measure of indu.stryperformance to be consistent with the compos-ite nature of our firm performance variable. Next,to control for the effects of industry tisk. weobtained the industry heta for each of the 96 indus-tries from Research Insight. Finally, we incorpo-rated information asymmetry into the analysis fortwo reasons. First, past data suggest the level ofinformation asymmetry in the environment mayimpact industry-level profitability (e.g., Nayyar.1990) as well as firm strategy (Nayyar. 1993: Nay-yar and Templeton. 1994). Second, the inclusionof information asymmetry helps account for anyeffects industry complexity/knowledge intensive-ness might have on human capital investments. Toassess information asymmetry, we used a four-itemscale that measured customers' abilities to ascer-tain the underlying quality of the service offeringas well as their ability to differentiate among theoutput of competing firms (alpha = 0.90). See theAppendix for a list of the specific informationasymmetry items.

Confirmatory factor analysis of all the self-reported constructs suggested fhat the modelprovided a good fit to the data. The CEA resultedin a chi-square statistic of 150 with 51 degreesof freedom. Since the chi-square was less thanthree times the degrees of freedom, a goodfit was implied (Carmines and Mclver. 1981).Furthermore, other measures of the fit indexes allexceeded the critical levels suggested by Bentlerand Bonett (1980) (comparative fit index ^ 0.983:goodness-of-fit index = 0.912: incremental fifindex = 0.983). Construct reliability was assessedby calculating Cronbach's alpha for each of theconstructs. All of the scales reached the 0.70a'suggested by Nunnally (1978). Moreover, thestandardized loadings of all measurement itemsto their respective constructs were significant atthe p < 0.05 level, suggesting that the scales forthe constructs had convergent validity (Montoya-Weiss, Massey. and Song. 2001). In addition,none of the confidence intervals of the phivalues contained a value of one ( ; J < O . O I ) ,

Copyrigtil John & Sons. Lid. Stnit. J.. 25: M5-W |2(H)4)

Strategic Positioning. Htwuin Capital, attd Pe/fonnance 91

again suggesting that the constructs possesseddiscriminant validity (Bagozzi and Phillips, 1982).

RESULTS

Strategic positioning and human capital

Correlations, means, and standard deviations ofall variables are shown in Table I. To test ourstrategic positioning-human capital hypotheses (1through 3), we used ordinary least squares regres-sion analysis. In doing so, we regressed the vari-able human capital on the three main effects (cus-tomer co-productiou. customer contact, and ser-vice customization) as well as on the two controlvariables (Jinn siz.e and information asymmetry).Results of this analysis are shown in Table 2.

Hypothesis I suggested a negative relation-ship between customer co-production and humancapital. The results support this hypothesis (h ——0.300. p < 0.01). This linding upholds our asser-tion that higher levels of customer co-productionact to standardize customer inputs into the pro-duction process, thereby teducing the need forhigher levels of human capital among a firm's pro-duction employees. This finding is also consistentwith those who make the corollary argument thatincreased co-production may allow firms to shiftproduction activities to custimiers.

Hypothesis 2 suggested a positive relationshipbetween customer contact and human capital. Theresults aiso confirm this contention (/j=; 0.194.p < 0.01). This finding supports the argument thatservice organizations with high levels of customercontact in the production and delivery processrequire greater levels of human capital in orderto handle the increased uncertainty brought about

Table 2. Results of regression analysis ofstrategic positioning on human capital

Strategic positioning

Si/.cInrormation asymmetryCttstoiiier co-prodtictionCttstomer contact.Service custotiii/ation

/?- = 0.24F ^ 12.401-

Utitiian capital

-0.086-0.123*-0.300"

0.194"0.212-

• / ' <-0.05; •• p < 0.01; *- /> < 0.001

Copyrighl tT 20U3 John Wik-y & Suns. LlJ.

by this interaction. As such, incteased investmentsin human capital are needed.

ln Hypothesis 3, we argued there would exista positive relationship between service customiza-tion and human capital. Again, the results supportthis contention (h = 0.212, p < 0.01). Thus, thereis strong evidence to suggest that service orga-nizations with adaptable production and deliveryprocesses are investing heavily in human capitalso that employees can process the correspondingvariance in customer demands.

Strategic positioning, human capital, andperformance

In order to test our hypotheses examining how the•fit' between strategic positioning and human cap-ital influences organizational performance (Hypo-theses 4. 5, and 6). we used moderated hier-archical regression analysis. In step I of thisanalysis, we entered the control variables (,v/,-(̂itulustry pctfortnance, itidustry heta. and itifonna-tion asytmuelry) and the main effects variables[customer co-production, customer coittitct. .ser-vice customiz.atiott. and human capital). Significanteffects here indicate direct telationships betweenthese variables and otganizational performance.This procedute eliminates any tiiain effects onperformance prior to examining potential strate-gic positioning-hutiiati capital interaction, or fit.effects (Stone and Hollenbeck, 1989). In step 2. thecross-products of each of the strategic positioningvariables and human capital (e.g., customer co-prodtiction x human capilal) were entered as a set.Entering the interaction terms simultaneously aidsin controlling for possible multicollinearity amongthe variables. Evidence of moderation exists whenthe set of interaction terms accounts for signif-icant residual variance in the dependent vari-able. A significant R' change here indicates thatstrategic positioning and human capital interactto influence performance. To understand the spe-cific relationships between the strategic position-ing-human capital interactions and performance,we then examined the individual interaction termsin the regression equations. Table 3 shows theresults of our moderated hierarchical regressionanalysis.

When examining the relationships among thestrategic positioning characteristics, human capital,and perfonnance. our regression model indicatesthat adding the strategic positioning and human

J.. 25: (2tXH)

92 B. C. Skaggs and M. Youndt

Table 3. Results of hierarchical regression analysis of strategic positiotiing,human capital, and perlnnnance

Variahles:

Control and main effectsSi/.eIndustry performanceIndustry betaInlonnalion asymmetryCtistomer co-prodtictionCustomer contaciServiee customizationHuman capital

Interaction efTectsCustomer co-production x Human capitalCustotner contact x Huinan capitalService customization x Human capitalInformation asymmetry x Human capilal

R-AFF

Step 1

0.0280.0070.0120.033

-0.0230.032

-0.176^^0.325***

0.107

2.59"

Step 2

0.0! 7-0 .0! !

0.0270.027

-0.0(i70.053

-O.!78*0.360***

-O.!45*-O.!7O*

0.383***-0.!60*

0.! 370.2447.718^"4.57"

0.05: *• 0.01: —/> <. 0.001

capita! interactions into our hierarchical retzressionanalysis in step 2 explains significant incremen-tal variance in pertbrmance (AR- — 0.137. AF =7.72, p <O.Ofll). Thus, there is strong supportfor the general proposition that strategic posi-tioning characteristics interact with human capitalto influence organizational performance. Further-more, each of the individual hypothesized inter-action terms (customer co-production x humancapital, cu.sttmier ctmtact x human capital, andsenice customizatioti x human capital) was sig-nificant.

As predicted in Hypothesis 4. customer co-producfion interacted with human capital to havea statistically significant, negative relationshipwith organizational performance lh = -0.145;p < 0.05). A plot of this interaction in Figure !(a)confirms our belief that higher performance willbe associated with service organizations thatcouple higher levels of customer co-productionwith decreased human capilal. On the otherhand., as these firms assume a larger portion ofservice production (i.e., as customer co-productiondecreases), increasing levels of human capital willbe linked tt) higher organizational performance.

The interaction between customer contact andhuman capital was also significantly related to per-formance (h = —0.170. p < 0.05). However, thedirection of this relationship was contrary to our

expectations stated in Hypothesis 5. The graphicaldepiction of this relationship in Figure l(b) revealsthat as customer contact increases, high levels ofhutiian capital are associated with decreasing per-formance, while low levels of human capital arelinked to performance increases.

The results for our final hypothesis. Hypothe-sis 6, confirmed our prediction that service cus-tomization and human capita! interact to affectperformance (h = 0.383. /? < 0.001). As expected.Figure l(c) illustrates that the combination of highlevels of customization and human capital arerelated to increased peiformance. Conversely, per-formance suffers substantially when increased ser-vice customization is coupled with low levels ofhuman capital.

DISCUSSION

Findings surrounding hypothe.sizedrelationships

Service organizations constitute almost SO per-cent of employment in the United States andover 75 percent of gross domestic product (UnitedStates Bureau of Labor Statistics, 1997). Yet, therehave been few studies that examine the effectsof customer participation in the service produc-tion and delivery process. Our study begins to

Cupyrjgtil L. 2m} Jolin Wiley & Snjis. l.liJ- Stiiit. .. 25: (21N)4)

(a)1 -1

0.5-

u. ot

0

-0.5 H

-1

(b)

Li. O

1 -1

0.5-

0

-0.5-

-1 -

Strategic Positioning, Human Capital, and Perfomiance 93

-2 . I X - - - 1 " *

Customer Co-production

Low Human CapitalHigh Human Capital

-2

Customer ContactLow Human CapitalHigh Human Capital

Service Customization

Low Human CapitalHigh Human Capilal

Figure I. Strategic positioning unJ htinian capilal inicractions

address this area by taking a customer interactionapproach to examining how certain strategic posi-tioning choices are related to human capital, andhow the interaction atnong strategic positioningand human capital influences organizational per-formance. We found sttong empirical support toindicate that strategic positioning (i.e.. customerco-production, customer contact, and service cus-tomization) is in fact related to human capital andthat the proper 'fit' among these variables is asso-ciated with changes in tirm performance.

To begin, we found that when organizationschoose to utilize more customer co-productionthey are less likely to make investmentsin human capital in their service productionprocesses. As hypothesized, this most likely resultsfrom organizations simplifying and standardizingtheir production processes as co~pioductionincreases. Such well-defined environments reduce

the cognitive demands placed upon employeesinvolved in the production process, whieh in turnreduces the organization's need to invest heavilyin developing and selecting human capital.

Our perfiirmance analysis with regard to the cus-tomer co-production and human capital interactionprovides some interesting insights. As expected.as co-production increases the performance ben-efit of high levels of human capital diminishessubstantially. Thus, on the surface it appears thatfirms are behaving rationally by coupling higherco-production with lower human capital (our resultfrom Hypothesis I). However, a detailed inter-action plot (Figure ia) shows that even thoughthere is a diminishing benefit to having highhuman capital when co-production increases, firmswith high huinan capital always outperform thosewith low human capital regardless of the level ofco-production. As such, under most co production

Copyrighl o 2003 John Wiley & Sims. LLJ. Stnii. Mv'i't. ./.. 25: H.S-W Ci

94 B. C. Skaggs and M. Youndt

circumstances organizatiotis would be wise toemploy high levels of human capital even thoughthe benefits of doing so arc substantially reducedunder conditions of moderate to high co production.

With regard to customer contact, we found aveiy .strong, positive link with human capital. Thisfinding implies that as customer-employee con-tact increases (i.e., as more of the productionprocess is exposed to customers), organizationsemploy higher levels of hutnan capital in orderto deal with the heightened information Hows andvariability resulting ftom this increased interac-tion. While such organizational decisions appearlogical and were consistent with our predictions,our intetaction analysis (Figure Ib) illustrates thatsuch a coupling can have detrimental perfonnanceconsequences. For example, combining low cus-tomer contact with low human capital resultsin extretiieiy low performance, whereas couplinghigh human capital with the same low contact leadsto veiy high performance. With regard to high cus-totiier contact, organizations exhibited vety similarperformance whether they utilized high or low lev-els of human capital when adopting this type ofstrategic position.

One explanation for our unexpected findingscan be found in an offshoot of agency theory.coined control theory (Snell, 1992). At its basiclevel, control theory outlines mechanisms organi-zations can use to affect the behavior of employ-ees in a manner that serves the best intetests ofthe firm. Assuming that managers of service firmsare interested in meeting the needs of customers.,then investments in hutnan capital through selec-tion and training can be used as a mechanism tohelp ensure that employees have the proper skillsto meet these needs. This approach to employeecontrol is referred to as input conlrol. Alterna-tively, firms can use hcliaviotrt! cotitfol by directlymonitoring their employees to ensute that they areacting in a manner consistent with meeting cus-tomer demands. We suggest that when customercontact is high, it is possible that customers eanserve as extra-organizational monitoring agents forservice firms. Rather than requiring managementto monitor employee behavior, it is the customersthemselves that ensure their needs are met throughdirect eotnmunication with ptoduction employ-ees. Due to this monitoring role by customersbrought about by a high degree of contact with theproduction ptocess, service organizations may beable to economize on their level of investments in

hurnan capital. Thus, firms employing high levelsof customer contact may be able to meet cus-tomer needs with either input control (i.e.. higherhuman capital) or behavioral control (i.e.. lowerhuman capital). This would explain why we foundlittle performance differences among firms pursu-ing high levels of customer contact. When cus-tomer contact is low. however, behavioral controlihrough customer tiionitoring may not be possibleas customers" interactions with production employ-ees are reduced. In this case, service firms usinginput control (i.e.. higher human capital) wouldmost likely achieve higher performance. Thoughthis explanation is consistent with our results, it isbut one possible reason for our findings. For exam-ple, it is also possible that employees were notgiven enough deeision-making discretion to dealwith the added uncertainty brought about by bigbcustomer contact, thereby reducing customer sat-isfaction and hence firm performance. Thus, muchmore research is needed to fully understand theeffects of customer contact on human capital.

We also found a strong, positive relationshipbetween service customization and human capi-tal. Not surprisingly, service otganizations employhigher levels of hutnan capital in their produc-tion processes when confronted by this poten-tial for substantial variance in customer demands.Under such a condition, employees need to ascer-tain a customer's specific needs and determinehow to provide customized services that meetthose idiosyncratic needs. Our interaction results(Figure lc) are consistent with this rationale andhighlight the importance of human capital to ser-vice customization. Organizations matching highhuman capital with highly adaptable service pro-duction exhibited extremely high levels of perfor-mance, wbile companies using low human capitalwith the satne strategic position experienced severeperformance problems. On the other hand, whenservice customization was low. organizations per-formed significantly better with low as opposed tohigh levels of human capital. As one might expect,it appears that service organizations with low cus-tomization tnay not be able to derive a requisiteamount of benefits to compensate for the additionalcosts associated with high levels of human capital.

Supplemental analyses

Although we did not hypothesize any direct rela-tionships between the strategic positioning chai-acteristics and performance, or between human

Copyright John Wtk-y & Sons, Ltd. Mgint. J.. 25: 85-W (2<K)4)

Strategic Positioning, Hutnan Capital, atui Performatice 95

capital and pertbrmance. our results provide someinteresting observations. The only strategic posi-tioning variable with a direct effect on perfor-mance was service customization (/; = —0.176./J < 0.05). This negative performance effect seemsto suggest that service customization, unless pack-aged and implemented correctly, may be verycostly and/or not valued by consumers. That is,service customization may have high ptoduetionand delivery costs without cotnmensurate pricingpower and therefore erodes profit margins unlesscoupled with the proper level of human capital.

The human capital variable itself had a verystrong, positive main effect on organizational per-formance {h = O..̂ 25, p < 0.001). This result lendsconsiderable support to the basic argument thathuman capital is vital to the production and deliv-ery processes of service organizations (e.g.. Mills.1986). Thus, all else being equal (e.g., strategicpositioning, industry environtnent). more hutnancapital appears to be better than less human capital.This result is also consistent with the visual plotsof the performance effects of the strategic posi-tioning-human capital interactions. While thesesinteractions revealed that certain strategic positionsrequired more or less human capital to maximizeperformance, the graphs illustrated that high lev-els of human capital were better than low levels inmost serviee production contexts.

In controlling for the potential effects of infor-mation asymmetry, we noticed that it too wassignificantly related to human capital.' Interest-ingly, the relationship was negative (/? = —0.123,p < 0.05). This is in direct contrast to what mostscholars have suggested, which is that industries

' We would like to thank an ani)nymous reviewer lor callingatiention lo this rekitionship iind encourajiing us lo investigate itfurther.

with high levels of information asymmetry atetypically characterized by complex, knowledge-intensive tasks and hence require high levels ofhutnan capital (Heskett. 1986: Normann. 1984).Because of the contrary nature of this finding, wedecided to investigate the effect of the interactionof information asymmetry and human capital onperformance. Our results indicated that the interac-tion was significant ih = —0.160. p < 0.05): how-ever, the graphical depiction of the relationshipwas again contrary to what one would expect basedon the literature. As highlighted in Figure 2. ser-vice lirms utilizing high levels of human capitalin environments characterized by low informationasymmetry had correspondingly high performance.while fittns employing low levels of human capitalin the same environments possessed low perfor-mance. At the other end of the spectrum, the per-fortnanee differences among firms taking either alow or high human capital approach narrowed sub-stantially when the environment was characterizedby high information asymmetry.

One explanation for these results is the poten-tial effect of information asytnmetry on the needfor organizational slack. Slack is typically con-ceptualized as excess resources that give firms agreater ability to respond to unforeseen demands(e.g., Thompson. 1967). In a service firm, invest-ments in human capital add to the level of orga-nizational slack, for they create a resource thatallows the finn to better cope with the potentialuncertainty customers bring to the service envi-ronment {Quinn, 1992). However, as informationasymmetry increases, customers are less able toascetiain the underlying quality of the serviceoffering and therefore have a reduced understand-ing as to whether or not their detnands are actu-ally being met (Nottnann. 1984). Such litnited

o -0.

-0.-0

8 n6 -4 -

2 -0 -2 ---4 -

6 -8 -1 -

t -

Information Asymmetry

Low Human Capital

High Human Capital

Figure 2. Inlbrmation asyinniLMrN and human capital interaction

Copyrighl '..• :(Hl.l John Wilcv & Sojn. Lid. .Sinit. J.. 2S; 85-49 (2(H)4j

96 B. C, Skaggs and M. Youndt

knowledge may therefore significantly decreasethe impact of a customer's demands upon theservice production process. In effect, informatiotiasymmetry may act as a buffer between the firmand its customers, allowing the firm to reduce itsinvestment in slack resources, which in this easeis human capital. In low information asymmetryenvironments, however, there is not much of aninformation buffer between the firm and its cus-tomers, thereby necessitating the use of high levelsof human capital. Much more research is needed tofully understand the impact of information asym-metry on service organizations.

Limitations and additional future directions

One limitation of this study involves the tim-ing of the effects between strategic positioningcharacteristics, investments in human capital, andperformance. Though we developed our customerinteraction arguments in terms of the impact ofstrategic positioning on human capital and theirsubsequent affect on organizational performance,other directional influences are possible. For exam-ple, it is reasonable to contend that successfulorganizations—regardless of the strategic posi-tioning characteristics they implement—have theresources necessary to invest in human capital.Likewise, it ean also be argued that high levels ofhuman capital afford organizations the ability topursue various strategic alternatives. On the otherhand, firms currently exhibiting low performancemay invest in human capital and certain strategicpositions in order to improve their performance.The cross-sectional nature of our study prohibitsus from excluding any of these possibilities. Futureresearch might look at strategic positioning charac-teristics, human capital, and performance over timeto examine the sequential and reciprocal relation-ships among these aspects of service organizations.

A second limitation involves the object of thestudy. Consistent with our customer interactionapproach, our study focused on the serviee produc-tion and delivery process. This focus allowed usto target a relatively homogeneous subpopulationof an organization's employees, thereby increas-ing the accuracy of our human capital measure-ments. However, doing so limits the generaliz-ability of our results to the production process.Future studies are needed to explore the linkagesbetween strategic positioning, human capital, andperfonnance in other subpopulations of employees

within service organizations. In a related matter,scholars may also consider examining these inter-relationships in diversified firms, as we limitedour study to nondiversified organizations primarilycompeting in a single industry.

A third limitation concems the relationshipbetween organizations' decision-making structuresand the level of human capital in their productionprocesses. We have argued that as variability incustomer demands increases, service organizationswill increase the level of human capital involved inthe production process to address this uncertainty.However, it is likely that these same productionemployees will also require higher degreesof autonomy so that they may address thisuncertainty. In the present study, we assumedthat managers would give the requisite amount ofdecision-making power to production employeesdepending on the level of variability theyencountered. If, however, managers did not grantsueh authority to production employees, thenany potential returns from investments in humancapital may not materialize. For example, asmentioned previously, our contrary performancefindings surrounding the customer contact-humancapital interaction may result from organizationsnot providing the requisite amount of decision-making authority to their production employees.This suggests there may exist both direct andinteraction effects among strategic positioning,human capital, and the decision-making structurethat ultimately impact perfonnance—effects thatwere not accounted for in the present study. Dueto the potential impact of decentralization on aservice firm's ability to handle uncertainty, weencourage future research in this area.

Besides these areas for future research, twoothers are of note. The first concerns the roleof technology in service organizations. As statedabove, our study focused on the human capitalrequirements surrounding the service productionprocess. Past research on production processes hasmostly examined the throughput aspect of produc-tion, in particular the use of technology in effi-ciency gains (e.g., Thompson, 1967; Woodward,1965). However, in the present study we didnot measure technology. Instead, we assumedit to be a multifaeeted resource that is appliedto achieve efficiency gains in a manner consis-tent with the organization's strategic positioning.Though this approach was theoretically consistent.

Copyrighl © 2(K)3 John Wiley & Sons. Ltd. J.. 25: 8.^-99 (2004)

Strategic Positiotiing, Hitman Capital, and Performance 97

it was nonetheless an oversimplification. Contin-uing advances in technology allow it to be usedin more difficult and complex tasks. Although ourresults show strong support for our argument thatservice firms use human capital to address vari-ability in customer demands, it is possible thattechnology is playing an ever-increasing role inallowing firms to contend with this uncertainty.That said, the reality is we know very little aboutthe role of technology in services. For instance,why is il that new technologies have brought forthvast productivity gains in the manufacturing sec-tor, while very few gains have occurred in theservices (United States Bureau of Labor Statistics.1997)? Another question involves the creation ofnew technology by service firms. Much of what aservice firm sells is a process. When such a firmdevelops a new process for delivering the service.this is by definition a new technology. However,processes are very difficult to patent. As such, newtechnologies in services may be easily copied bycompcti(t)rs. which in turn may reduce these orga-nizations" desire to invest in innovation. Thesequestions have only scratched the surface of poten-tial topics involving technology in the services. Agreat deal more research is needed in this area.

The second area involves the skill level of cus-tomers. In the present study we assumed that allcustomers possessed a similar level of skill. How-ever, it is likely that differences do exist amongcustomers based on their skill and/or knowledgesurrounding the service. For example, Nayyar andTempleton (1994) suggest that expert buyers mayhave a better familiarity with the service thanwould novice clients. As such, expert customersmay be able lo perform more complex tasks dur-ing service production, thereby reducing the needfor higher levels of human capital, ln addition,customer skill level could also impact the strate-gic options available to service firms. When cus-tomer skill level is very low, firms may encounterdifficulty in adopting a production process highin customer co-production (i.e.. one that requiresmuch effort on the part of customers). Finally,the level of customer skill may also impact abuyer's susceptibility to the effects of informa-tion asymmetry (Nayyar and Templeton. 1994).For example, highly skilled customers may havea better understanding of what constitutes qualityin the service offering. As such, this could haveimplications for the level of slaek resources in theform of human capital that service firms require.

These are just a few examples of how customerskill level can affect service organizations. Again,much more research is needed.

CONCLUSION

Previous research on production and human cap-ital has typically focused on the nature of thetask. However, this body of research implicitlyassumes a separation between production and theenvironment. Recent works have suggested thatin service environments a client-firm interactionoccurs that can create higher levels of uncertaintyfor the tirm. As a result of this interaction, ser-vice organizations must adapt elements of theirproduction processes to address this uncertainty.The current study extends this idea by suggestingthat the strategic positioning of service productiondetermines the level of uncertainty arising from theclient-firm interaction, and hence the human capi-tal required to handle this uncertainty. In addition,we found partial support for performance differ-ences among serviee firms as a result of their fitbetween strategic positioning and human capital.

Thus, the findings in this study add to the grow-ing belief among researchers that a different setof dynamics may be at work in serviee organiza-tions. With this sector of the economy account-ing for the vast majority of both employment andgross domestic product in the United States, webelieve that more research in this area is war-ranted. We hope that the findings presented herewill encourage researchers to continue investigat-ing the dynamics driving these organizations.

ACKNOWLEDGEMENTS

The authors would like to thank Jason Shaw for hishelpful comments on an earlier draft. We wouldaiso like to thank two anonymous reviewers fortheir significant contributions to the developmentof this paper.

REFERENCES

Aiken LS. West SG. 1991. Multiple Regre.s.sion: Testingatul Interpretin^i Interactions. Sage: Newbury Park.CA.

(.• 2(H11 Jnhn Wiley & Sons. Lid. Mamt. J.. 25; K.'i-W 12(X)4)

98 B. C. Skaggs and M. Youndt

Argolf L. 1982. Isiptil tiiiccrlainly and organizationatcoordination in hospital emergency units. Administra-tive Science Quarterly 27: 420-434.

Arthur JB. 1994. Effecls ol" human resotircc systems onmanufacturing perfomiiincc and turnover. Academy ofManagement Journal 37: 670-687.

Bagozzi RP. Phillips LW. 1982. Representing andtesting organizational theories: a holistic construct.Administrative Science Quarterly 27: 459-489.

Becker GS. 1964. Human Capital: A Theoreticaland Empirical Analysis, with Special Reference toEducation. Columbia University Press: New York.

Bentler PM. Bonett DC 1980. Significance tests andgoodness of fit tests in the analysis of covariancestructures. P.sychological Bulletin 88: 588-606.

Bharadwaj SG. Varadarajan PR. Fahy J. 1993. Sustain-able competitive advantage in service industries: aconceptual model and research propositions. Journalof Marketing 57: 83-99.

Boxall P. Steeneveld M. 1999. Human resource strategyand competitive advantage: a longitudinal study o!'engineering consultancies. Journal of ManagementStudies 36: 442-463.

Brush TH. Art/. KW. 1999. Toward a contingentresource-based theory: the impact of informationasymmetry on the value of capabilities in veterinarymedicine. Strategic Management Journal 20(3):223-250.

Carmines EG, Mclver JP. 1981. Analy/.ing modelswith unobserved variables: analysis of covariancestructures. In Social Measurement: CurrenI Issues.Bohrnstedt GW. Borgatta EF (eds). Sage: NewburyPark. CA: 65-115.

Delery JE. Doty DH. 1996. Modes of theorizingin strategic human resource management: testof universali.stic. contingency, and conligurationalperformance predictions. Academy of ManagetnentJournal 39: 802-835.

Heskett JL. 1986. Managing in the Sen-ice Economy.Harvard Business School Press: Boston. MA.

Hitt MA. Bierman L. Shimizu K. Kochhar R. 2001.Direct and moderating effects of human capital onstrategy and performance in professional servicelirms: a resource-based perspective. Academy ofManagement Journal 44: 13-28.

Huselid MA. 1995. The impact of human resourcemanagement practices on turnover, productivity.and corporate financial perfonnance. Academy ofManagement Jounud 38: 635-672.

Jones GR. 1987. Organi/ation-clienl transactions andorganizational govemance structures. Academy ofManagement Journal 30: 197-219.

Keats BW. Hiti MA. 1988. A causal model of linkagesamong environmental dimensions, macro organiza-tional characteristics, and performance. Academy ofMatiagement Journal 31: 570-598.

Kotler P. 1983. Principles of Marketing. Prentice-Hall:Englewood Cliffs. NJ.

Lovelock CH, Yip GS. 1996. Developing global strate-gies for service businesses. California ManagementReview 38: 64-86.

MacDuffie JP. 1995. Human resource bundles andmanufacturing performance: organizational logic andHcxiblc production systems in the world autoindustry. Industrial and Labor Relations Review 48:197-221.

Maister DH, Lovelock CH. 1982. Managing facilitatorservices. Sloan Management Review 23: 19-31.

Mills PK. 1986. Managing Serxice Industries: Orga-nizational Practices in a Postitulustrial Ecotwmy.Ballinger: Cambridge, MA.

Mills PK. HallJL, LeideckerJK. Margulies N. 1983.Flexiform: a model of professional service orga-nizations. Acadetny of Managetnetit Review 8:118-131.

Mills PK. Moberg D. 1982. Perspectives on the technol-ogy of service operations. Academy of ManagementReview 7: 467-478.

Mills PK. Morris JH. 1986. Clients as 'partiar employeesof service organizations: role development in clientparticipation. Academy of Management Review 11:726-735.

Montoya-Weiss MM, Massey AP. Song M. 2001. Get-ting it together: temporal coordination and conflictmanagement in global virtual teams. Academy of Man-agement Journal 44: 1251-1262.

Nayyar PR. 1990. Information asymmetries: a source ofcompetitive advantage for diversilied service firms.Strategic Management Jounud 11|7): 513-519.

Nayyar PR. 1993. Performance effects of informationasymmetry and economics of scope in diversifiedservice firms. Academx of Managemetit Jounud 36:28-57.

Nayyar PR. Templeton PL. 1994. Seller beware: informa-tion asymmetry and the choice of generic competitivestrategies for service businesses. In Advances in Ser-vices Marketing and Management. Vol. 3. Swartz TA.Bowen DE. Brown SW (eds). JAI Press: Greenwich,CT: 95-126.

Normann R. 1984. Service Managetnent: Strategy am!Leadership in Senice Busine.s.ses. Wiley: Chichester.U.K.

Nunnally J. 1978. P.nrhomeiric Theoiy (2nd edn).McGraw-Hill: New York.

Pfeffer J. 1994. Competitive Advatnage through People:Unleashing the Power of the Work Eorce. HarvardBusiness School Press: Boston. MA.

Porter ME. 1980. Competitive Strategy: Tecluiiques forAnalyzing Industries and Competitors. Free Press:New' York.

Porter ME. 1996. What is strategy? Harvard BusinessReview 74(6): 61-78.

Quinn JB. 1992. Intelligent Enterprise: A Knowledge atidService Based Paradigm for lndustr\-. Free Press: NewYork.

Schultz TW. 1971. Investtnent in Human Capital: TheRole of Education and Research. Free Press: NewYork.

Scott WR. 1992. Organizations: Rati(mat. Natural, andOpen Systems (3rd edn). Prentice-Hall: EnglewoodCliffs. NJ.

Shostack GL. 1987. Service positioning through struc-tural change. J<mrnal of Marketing 51: 34-43.

Copyright © imi' John Wik-y & Sous. Ltd. Sirui. J-. 25; 85-9^ (2004)

Strategic Positioning, Human Capital, and Performance 99

Simon JL. Burstcin P. 1985. lia.sic Research Methods inSocial Science (3rd edn). ML-Graw-Hill: New Y(.irk.

Snell SA. 1992. ConlntI theory in siralcgic humanresource managemeni: the mediating cfTect ofadministrative informalion. Acadeniy of Mdncij^emenlJournal 35; 292-327.

Snell SA. Dean JW Jr. 1992. Integrated manufacturingand human resources management: a human capilalperspective. Academv of Management Journal 35:467-504.

Stone EF. Hollenbeck JR. mS9. Clarifying some coti-troversial issues surrounding stalistical procedures ofdetecting moderator variables: empirical evidence andrelated matters. Journal of Applied Ps\clwlog\ 74:3-10.

Tansik DA. 1990. Balance in service systems design.Journal of Busine.s.s Research 20: 53-61.

Thompson JD. 1967. Ori^anizations in Action. McGraw-Hill: New York.

Tushman M. Nadier D. 1978. Information processingas an integrating concept in organizational design.Academy of Manaf^emenl Review 3: 613-624.

United States Bureau of Labor Statistics. 1997. MonthlyiMbor Review 119(3).

Upah GD. 1980. Mass marketing in service retailing: areview and synthesis of major methods. Journal ofRelailinji 56: 18-28.

Wathen S. 1995. Designing services: an information-processing approach. Inreniational Journal of ServiceInduslry Manaf^ement 6: 64-76.

Williamson OE. 1979. Transaction-cost economics: thegovernance of contractual relations. Journal of Lawand Economics 22: 232-261.

Woodward J. 1965, Induslrial Ori^ani~ari(m: Tlteorx andPractice. Oxford University Press: London.

Youtult MA. Snell SA, Dean JW Jr. Lepak DP. 1996.Human resource management, manufacturing strategy,and firm performance. Academv of ManagemeniJournal 39: 836-866.

APPENDIX: QUESTIONNAIRE ITEMS

Customer co-production

Relative ta competitors, anr firm:

1. requires customers to perform functions similartt) Ihat of an employee.

2. requires customers to become heavily involvedin producing the service.

3. "minimizes ihe amount of time customers spetidin Ihe service production process.

4. *pertbnns many tasks for customers (e.g., rentalcar company offering free pick-up and drop-offservice).

5. *is conveniently located near customers.

Customer contact

Relative to competitors, most of our firm's.service production:

1. "occurs out of customers' view (i.e.. in the'back office").

2. occurs in full view of customers (i.e.. in the•front office').

Service customization

Relative to competitors, onr firm:

1. changes how our service is offered for eachcustomer.

2. *offers a service which is similar from customerto customer.

3. allows our customers to dictate how the serviceis offered.

4. "^pertbrms the same procedures for each cus-tomer.

5. requires a great deal of information fioni eachcustomer before producing the service.

Human capital

Relative to compelilors, our firm:

\. spends more money per employee on trainitig.2. spends more hours per year training employees.3. hires etnployees with high levels of prior expe-

rience.4. hires employees with high levels of prior train-

ing.5. hires employees with high levels of education.

Information asymmetry

Relative to other industries, how accurately do thefollowing describe the service offered inyour industry'/

1. It is difficult for customers to understand howservices are actually produced.

2. Customers cannot determine firms' level ofeffort in service production.

3. It is difficult for customers to make comparisonsof the service offering across competing firms.

4. It is difficult for customers to determine thequality of the service offering prior to purchase.

All items were on a scale of I (not accurate) to 7(very accurate).

*Indicates item was reverse coded.

Copyright © 2003 John Wiley & Sons. Lld- Sirai. Mgml. J.. 25: 85-'W (2(XW)