Ee

Sa

b

c

a

ARRAA

KEFIPPPST

1

eotpb2piWaetp

(

0h

Journal of Operations Management 30 (2012) 509–520

Contents lists available at SciVerse ScienceDirect

Journal of Operations Management

jo ur nal home page: www.elsev ier .com/ locate / jom

ffect of purchase volume flexibility and purchase mix flexibility on-procurement performance: An analysis of two perspectives

arv Devaraja,∗, Ganesh Vaidyanathanb, Abhay Nath Mishrac

Management Department, University of Notre Dame, Notre Dame, IN 46556, United StatesSchool of Business and Economics, Indiana University South Bend, South Bend, IN 46634, United StatesRobinson College of Business, Georgia State University, Atlanta, GA 30302-3989, United States

r t i c l e i n f o

rticle history:eceived 14 April 2011eceived in revised form 12 July 2012ccepted 8 August 2012vailable online 23 August 2012

eywords:-procurementrequencynformation sharingurchase volume flexibilityurchase mix flexibilityerformanceupplier customization

a b s t r a c t

Despite the widespread adoption of e-procurement by firms in recent years, academic research exam-ining the mechanisms through which e-procurement applications lead to performance has been scarce.Anecdotal evidence points to numerous situations where companies have failed to harness the poten-tial of e-procurement. In this paper, we argue that online purchase volume and mix flexibilities facilitatedby these applications play a significant role in the ability of firms to benefit from e-procurement. Weexamine this tenet from both an economic as well as a social perspective. We propose that increasedonline purchase volume flexibility as well as online purchase mix flexibility can be facilitated by twomechanisms – supplier customization as explained by transaction costs perspective, and informationsharing between supply chain partners using a social exchange theoretical perspective. The increasedpurchase volume and mix flexibility in turn leads to better performance along the dimensions of cost,quality, and delivery. We present and test a nuanced perspective where we argue that (i) the effect ofsupplier customization on both purchase volume and mix flexibilities will be moderated by the frequency

rustof transactions conducted online, and (ii) the effect of information sharing on both purchase volume andmix flexibilities will be moderated by trust in the supplier. We estimate our research model using surveydata collected from 130 purchasing and procurement managers. We find strong support for our proposedresearch model with results indicating that purchase volume and mix flexibilities play a vital mediatingrole in impacting e-procurement performance. Theoretical and practical implications of the findings are

discussed.

. Introduction

The procurement function is considered the focal point offfective strategic supply chain management in contemporaryrganizations (Talluri and Sarkis, 2002). Advances in informationechnology (IT) have enabled firms to streamline the procurementrocess and value-chain activities to achieve substantial economicenefits (Devaraj et al., 2007; Gosain et al., 2005; Hill and Scudder,002). In particular, e-procurement – the automation of a firm’srocurement process – is becoming increasingly recognized for

ts potential to improve business operations and reduce expenses.hile a number of firms have been able to leverage e-procurement

pplications to lower costs and increase overall profitability, oth-

rs have grappled with the intricacies of these technologies andhe mechanisms through which they can impact procurementractices and economic benefits (Olson and Boyer, 2003). In this

∗ Corresponding author.E-mail addresses: sdevaraj@nd.edu (S. Devaraj), gvaidyan@iusb.edu

G. Vaidyanathan), amishra@gsu.edu (A.N. Mishra).

272-6963/$ – see front matter © 2012 Elsevier B.V. All rights reserved.ttp://dx.doi.org/10.1016/j.jom.2012.08.001

© 2012 Elsevier B.V. All rights reserved.

paper, our focus is to examine how e-procurement influences pur-chase volume flexibility and purchase mix flexibility in relationto the traditional purchasing context, and how such flexibilitiesimpact e-procurement performance. The key roles played by e-procurement applications in enabling firms to enhance purchasevolume and mix flexibilities, and the consequent impact of suchenhancements, have been largely ignored in the existing literature.

In general, flexibility in business processes has become increas-ingly important for firms to respond efficiently to evolvingcustomer requirements, intense global competition, and rapidtechnological advancements (Sambamurthy et al., 2003; Swaffordet al., 2006). Thus, flexibility has been a topic of intense inter-est among operations management (OM) researchers (De Groote,1994; Gerwin, 1993; Jordan and Graves, 1995; Koste and Malhotra,1999; Suarez et al., 1996). However, prior research has focusedpredominantly on examining the different dimensions of manufac-turing flexibility, the relationship between these dimensions, and

the impact of these dimensions of flexibility on performance (Caoand Dowlatshahi, 2005). Flexibility in other business processes andthe role that information technology can play in obtaining flexibil-ity has not been investigated in detail in extant literature.

5 tions

tgaflai(vbtirtsft

boiflweOect(1ktpiob2ofcccflmWpRp

tiirasotteflmomplp

10 S. Devaraj et al. / Journal of Opera

We conceptualize purchase volume and purchase mix flexibili-ies as the ability of a firm to modify the mix and volume of inputoods dynamically using e-procurement with little penalty in timend effort. We draw upon Duclos et al. (2003), who suggest thatexibility allows managers to change product mix and volume, rel-tively quickly and without much struggle. Grounding our researchn the theoretical perspectives of transactions costs economicsTCE) and social exchange theory (SET), we propose that purchaseolume and mix flexibilities in e-procurement can be facilitatedy two major factors – information sharing and supplier cus-omization. We further propose that two theoretically-groundedmportant contextual factors – trust and frequency – accentuate theelationships between purchase volume and mix flexibilities andheir antecedents. This study provides a hitherto unexplored per-pective on purchase volume flexibility and purchase mix flexibilityacilitated by e-procurement and the antecedents that facilitatehem.

The enhanced purchase volume and mix flexibilities engenderedy e-procurement applications can also have substantial impactsn procurement process performance. While the prior literaturen manufacturing strategy suggests that cost, quality, delivery, andexibility are main competitive priorities and performance metrics,e suggest that flexibility may play a much more important role in

-procurement than has been acknowledged in the prior research.ur proposition is that other metrics – cost, quality and deliv-ry – are driven by online volume and mix flexibilities in theontext of procurement. While there is precedence in the litera-ure of researchers analyzing firm performance based on flexibilityDevaraj et al., 2007; Jack and Raturi, 2002; Kekre and Srinivasan,990; Suarez et al., 1996; Vickery et al., 1999), to the best of ournowledge, this is one of the first studies in OM that theorizeshe role played by online purchase volume flexibility and onlineurchase mix flexibility in the context of e-procurement and empir-

cally validates it. Prior studies have examined the role of qualityn e-procurement satisfaction (Vaidyanathan and Devaraj, 2008),uyer-side competence in B2B commerce (Rosenzweig and Roth,007), and organizational collaboration (Sanders, 2007). However,ur focus is on online purchase volume and mix flexibilities, andurthermore we propose that the use of flexible purchase pro-esses facilitated by e-procurement can enable organizations toope with changes in the marketplace to respond proactively to theontingencies presented by uncertain business environments. Suchexibility in procurement process may result in significant perfor-ance benefits along cost, quality and responsiveness dimensions.e test our research model with survey data collected from 130

urchasing and procurement managers from various industries.esults obtained from employing structural equation modelingrovide strong support for our research model.

This paper makes four key contributions to the existing litera-ure. First, we investigate the role of volume and mix flexibilitiesn e-procurement, which has been not been studied extensivelyn the existing literature. In fact, scholars have issued severalesearch calls for a fine-grained understanding of purchase volumend mix flexibilities (Duclos et al., 2003). This study responds touch calls. Our research model provides a nuanced understandingf the key roles played by purchase volume and mix flexibili-ies on e-procurement performance and the facilitating conditionshat enable firms to leverage these flexibilities in the context of-procurement. We demonstrate that purchase volume and mixexibilities can indeed impact other business process performanceetrics such as cost, quality, and delivery. Second, given that

rganizational understanding of successful e-procurement use, the

echanisms through which e-procurement use impacts business

rocesses, and the implications of such use are still lacking in theiterature (Bradley, 2005; Vaidyanathan and Devaraj, 2008) we pro-ose that e-procurement can be used as a means for firms to gain

Management 30 (2012) 509–520

flexibility. To the best of our knowledge, the flexibility aspect ofe-procurement has been largely ignored in the existing literature.This study provides a potential new direction for the examinationof e-procurement applications in firms. Third, we provide supportfor and extend the literature that suggests that the use of infor-mation systems can enhance the flexibility and agility of firms andenable them to respond to market signals with ease (Sambamurthyet al., 2003). Finally, by drawing on TCE and SET, this paper pro-poses, and finds empirical support for, two key enablers of purchasevolume and mix flexibilities, as also moderators that impact theinter-relationship between them.

The rest of the paper is organized as follows. We present the the-oretical background along with the hypotheses in the next section.In Section 3, we present the details of our empirical approach, anddiscuss the adequacy of the measures used in the study. The resultsof our analyses and findings of our study are discussed in Section4. Finally, in Section 5, we discuss the limitations and implicationsof this study, and suggest areas of further research.

2. Theory and hypotheses development

We first present our research model (in Fig. 1) as a roadmap,following which we discuss the theoretical underpinnings of themodel. As illustrated in Fig. 1, we suggest that online purchasevolume and mix flexibilities are key constructs that determinethe performance of the e-procurement process in organizations.Drawing on two different theoretical perspectives, we propose thatonline purchase volume flexibility and online purchase mix flexibil-ity can have two key antecedents. According to TCE, asset specificityis a key driver of transaction costs and the relationship betweensupply chain partners. Asset specificity has emerged as a reliableand consistent predictor of inter-organizational relationships andsourcing decisions (Poppo and Zenger, 1998). Asset specificity isrelated to alternative uses of the asset involved in the transaction,and it is measured by the lack of standardization. In our context,asset specificity is operationalized as supplier customization. Theextent to which a supplier has customized its assets, processesand tools to meet a buyer’s requirements is a significant indica-tor of its commitment to meet the idiosyncratic requirements ofthe buyer, and is a key economic factor that may impact onlinepurchase volume and mix flexibilities.

According to SET, pure economic factors are insufficientto explain inter-organizational relationships, and social factorsinform why firms engage in extensive coordination in the inter-ests of maintaining long term relationships. In our context, wehave conceptualized information sharing as the key social factorthat may influence online purchase volume and mix flexibili-ties. Scholars have acknowledged communication as an essentialfactor in e-procurement and suggested that product customiza-tion and information sharing constitute key characteristics ine-procurement processes (Chen et al., 2004; Doney and Cannon,1997; Muffato and Payaro, 2004; Subramani, 2004). Doney andCannon (1997) argue that supplier customization, and confidentialinformation sharing are among the two most important character-istics in inter-organizational relationships and the coordination ofactivities.

This paper has also conceptualized two moderating variablesgrounded in TCE and SET – frequency of transactions and trust – thatare likely to impact the direct relationships between supplier cus-tomization and purchase volume and mix flexibilities and betweeninformation sharing and purchase volume and mix flexibilities

respectively. The dependent constructs of interest in our model arecost, quality, and responsiveness obtained through online purchas-ing in comparison to traditional modes of purchasing. We discussspecific hypotheses related to buyer–supplier characteristics,

S. Devaraj et al. / Journal of Operations Management 30 (2012) 509–520 511

searc

pp

2

tdtpsctFtavts

icaisfiacSeectipeoi

tanOfiiT

Fig. 1. Re

urchase volume and mix flexibilities and procurement processerformance next.

.1. Transaction cost economics (TCE) and supplier customization

Supplier customization refers to the dedicated assets, such asools and equipments, and tailored processes that a supplier areevotes to the buyer firm, and is indicative of the willingness ofhe supplier to conform to the needs of the buyer firm. For exam-le, supplier customization includes initiatives on the part of theupplier to acquire specialized equipments in order to manufactureustomized and tailored products and to modify internal processeso conform to the needs of the buyer (Doney and Cannon, 1997).rom a transaction costs economics perspective, such customiza-ion initiatives constitute asset specificity. Because procedures,ssets and tools are customized to a specific buyer, their economicalue and salvage potential is considerably less if they are switchedo alternative transactions. Thus asset specificity tends to promoteupplier–buyer relationship on economic grounds.

From the buyer firm’s standpoint, supplier customization allowst to modify the supply of products rapidly in response to changingustomer needs because the dedicated resources of the supplier canccommodate buyer requests by modifying processes and employ-ng customized tools and assets. Environmental changes, includinghifting customer needs, can create significant uncertainty for arm. To mitigate this uncertainty, buyers minimize customizationt their end (Sia et al., 2008), and rather rely on suppliers with highustomization to facilitate product volume and mix flexibilities.uch accommodation by suppliers is common. For instance, Hsieht al. (2008) claim that market-oriented suppliers from Taiwanmphasize on the customization of their processes to satisfy theirustomers. Such a customization strategy requires a supply chainhat is both volume and mix flexible (Salvador et al., 2007). This maynclude flexibility to increase or decrease the volume of customizedroducts or change the mix of customized products. In consumerlectronics sector for example, buyers welcome flexibility becausef their need for high level of customization and frequent changesn product mix and delivery.

Supplier arrangements and customizations enable buyer firmso appropriately structure inter-organizational information flowsnd coordinate procurement processes, thereby reducing the effortecessary for resolving external uncertainties (Gosain et al., 2005).

nce supplier arrangements and customizations are in place, buyerrms can use e-procurement applications effectively to procure

nput materials that would serve customer requirements better.he rich communication enabled by e-procurement applications

h model.

allows firms to transmit features, specifications and requirementsmuch more succinctly and accurately to suppliers than would bepossible through traditional means, such as face-to-face meetings,paper documents and phone calls, and still convey all the require-ments so suppliers can manufacture the component. Additionally,suppliers’ customization levels influence the ability of buyer firmsto use technology in the entire procurement process, and conductbusiness electronically.

Due to reasons outlined above – lesser opportunistic behavior,idiosyncratic assets losing value unless relationships are contin-ued, evidence that the supplier firm cares for the relationship, – itis reasonable to believe that if a supplier firm engages in signifi-cant customization, the focal firm (buyer firm) will leverage on thisidiosyncratic investment to gain more flexibility from the relation-ship than in situations where the supplier amasses general purposeassets and exploits them to supply several other buyers.

Based on the above discussion, we hypothesize

H1. Supplier customization (asset specificity) will be positivelyassociated with online purchase volume and mix flexibilities.

In TCE, frequency is the notion of how often the parties trans-act with each other and is aimed at capturing the significance orimportance of these transactions (Teo and Yu, 2005). The followingperspectives highlight the moderating role that frequency of trans-actions has on the relationship between supplier customization andflexibility performance:

(a) Frequency translates to recency of information – The more fre-quent the transactions, the more recent will be the informationon volume and mix. With a customized setup and more recentinformation, the supplier will be better positioned to deliveron volume and mix flexibility. For instance, ceteris paribus, thequantity and mix of input materials a firm procures can be mod-ified more easily if the materials are procured with a higherfrequency because of the recency of information. As the needfor dynamic adjustment is more acute for firms that transactfrequently with their suppliers, it is evident that the impact ofsupplier customization on purchase volume and mix flexibilitywill be higher for such firms.

(b) Familiarity breeds success – Given that the supplier has alreadyinvested significantly in customized assets to service the focalcustomer, the more frequent exchanges between the supplier

and customer imply a level of familiarity and trust that enablesthe supplier to provide a level of service (in this case flexibility involume and mix) that would not have been possible otherwise.Frequent transactions indicate the need of a closer relationship

5 tions

(

bs

Hmtfbfl

2

nretmos1tsmsusatit(mi

liwbic2ttwp

12 S. Devaraj et al. / Journal of Opera

with suppliers (García-Dastugue and Lambert, 2003). Closerelationships between buyers and suppliers lead to better prac-tices in business processes

(c) Frequency as a proxy for significance of the relationship – Fre-quent transactions between a supplier and a customer mightdenote that in the supplier–customer dyad, each of them is asignificant trading partner for the other. From the supplier’sstandpoint, the customer being a significant trading partner andthe fact that the supplier has the ability to supply customizedservice provide a catalyst for the supplier to offer higher degreeof flexibility to the customer (the focal firm) who is a significanttrading partner.

d) Finally, frequency may be used as a basis to counter fluctuations involume and mix faced by the buyer firm. This argument impliesthat the buyer firm by interacting more frequently with thesupplier meets its own demand uncertainties. Further, the sup-plier’s customization allows the seller firm to accommodatesuch fluctuations.

Based on these reasons, frequency moderates the relationshipetween supplier customization and flexibility. Thus, we hypothe-ize:

2. The effect of supplier customization on purchase volume andix flexibilities will be stronger when the frequency of transac-

ions conducted between the exchange partners is high. That is,requency of transactions positively moderates the relationshipetween supplier customization and purchase volume and mixexibilities.

.2. Social exchange theory and information sharing

In contrast to the economic rationale of transaction cost eco-omics is the body of literature compiled by social exchangeesearchers that presents various social rationales for whyxchange partners might continue to sustain a relationship. Fromhis theoretical perspective, exchange partners will engage in infor-

ation sharing if this helps to foster their relationship for a varietyf reasons. First, attachment or prior history of the relationship is aignificant driver of commitment to a relationship (Seabright et al.,992). If two firms have had a long history of successful collabora-ion, they are more likely to engage in activities such as informationharing that benefit the relationship. Second, shared values ofacro-culture among the exchange partners are an important

ocial dimension (Young-Ybarra and Wiersema, 1999). These val-es denote a common understanding of not only the inter-firmetting but also about resolution of conflicts and disputes if theyrise. Third, reputation is an important attribute in exchange rela-ionships (Kollock, 1994). A firm is more willing to rely on and sharenformation with a partner that has a high reputation in the indus-ry. Reputation, in a sense, is a safeguard against deceptive behaviorParkhe, 1993) and thus firms are likely to transact and share infor-

ation more freely when the reputation of the exchange partnerss high.

Information transfer in a supply chain is defined as the regu-ated flow of information from one unit (e.g., firm, work group, orndividual) to another. In a systematic procurement arrangement

here regular transactions are expected to occur on a long-termasis via negotiated contracts with qualified suppliers, organized

nformation transfer between the trading partners may be cru-ial to achieve efficiency in recurring transactions (Kim et al.,005). Johnson et al. (2007) also emphasize the need of informa-

ion exchange by integrating e-business technologies. Accordingo them, deploying such technologies with purchasing teams toork effectively with critical supply chain partners provides com-etitive advantage to organizations. Inter- and intra-organizational

Management 30 (2012) 509–520

collaboration is shown to have a strong direct impact on organiza-tional performance (Sanders, 2007).

Organizations use their sense and respond capabilities toacquire more information about the potential changes and oppor-tunities and to respond to those possibilities. The informationsharing capabilities include the integration of the systems, deci-sions, and processes (Hsu et al., 2008). The availability ofinformation allows firms to sense the need for change in their cur-rent process configuration and to develop mechanisms for dealingwith change. Information sharing is needed to allow an enter-prise to sense the needs of the partners and communicate its ownneeds to them. Knowledge obtained through sharing of informationenables firms to react to unanticipated change and to be continuallyattuned to change. Firms that exchange information with suppliersextensively can respond to demand changes with agility. Frequentinformation exchange between a firm and its suppliers encouragesthe firm to share both tactical information (e.g., engineering changeorders, reject rates, inventory positions), and strategic information(e.g., product roadmap, demand forecast, cost curves) with suppli-ers. Such detailed information sharing enables the buyer and itssuppliers to coordinate their design efforts, production plans andshipping schedules more effectively.

Changes in demands give rise to uncertainty among suppliersand customers. To resolve this uncertainty, suppliers and cus-tomers need to share information about customer demands, andsupplier capacities and schedules. Supply chain coordination suf-fers when decision makers have incomplete information aboutlatest customer demands (Sahin and Robinson, 2002, 2005). Incontrast, Cachon and Fisher (2000) have found that the sharingof real-time demand information provides significant operationalimprovements. In addition to providing operational savings, firmshave come to realize that sharing information with business part-ners can facilitate meeting customers’ needs in a timely manner.The exchange of information with suppliers enables buyer firmsto communicate customer requirements to suppliers, and aidsin procuring parts and components that are required to sat-isfy changing customer requirements. As customers’ needs evolvedynamically, firms can gather information about them, share thisinformation with suppliers and respond to such events fasterthrough the use of e-procurement applications in comparison totraditional modes. Firms that share extensive information withsuppliers are likely to leverage their existing supplier base andexploit e-procurement solutions for enhanced collaboration, flex-ibility and transactional efficiencies (Jap and Mohr, 2002). Thesefirms are more open to using e-procurement solutions for coor-dinating production plans and shipping schedules on the Internetwith extant suppliers, and to transacting online (Premkumar andRamamurthy, 1995). The implementation of an e-procurementapplication makes it possible to digitalize delivery plans and shareextensive information with suppliers, which lead to greater flexi-bility and better control of the goods supplied (Muffato and Payaro(2004). Based on these arguments, we hypothesize,

H3. Information sharing facilitated by e-procurement will be pos-itively associated with purchase volume and mix flexibilities.

Young-Ybarra and Wiersema (1999) state that trust betweenorganizations will have a positive impact on the desire and abilityof partners to adjust to changing environmental demands. Higherlevels of trust in business partners indicate that an organizationbelieves in the competence or expertise of its partners to deliver.The trusting organization also expects that the trusted party willbe benevolent and not take advantage of the new conditions by

willful exploitation (Young-Ybarra and Wiersema, 1999). Trust isconsidered to be the central tenet of interorganizational exchanges,and has been posited to provide outcomes that are superior tothe ones that firms can obtain if they act solely in self-interest

tions M

(tdah

tmtmwotbaHmceb

Hflptp

2fl

ocihc2Umcandt(ei

nfloitflc

aattprono

S. Devaraj et al. / Journal of Opera

Anderson and Narus, 1990). Additionally, while increased informa-ion sharing enables suppliers and buyers to be flexible to customeremands (Rosenzweig et al., 2003), Handfield and Bechtel (2002)rgue that firms are more willing to be responsive to partners whoave exhibited trustworthiness.

The trust between exchange partners can strengthen the rela-ionship between information sharing and purchase volume and

ix flexibility through many reinforcing impacts. Higher levels ofrust encourage organizations to share accurate and frequent infor-

ation with partners because of the belief that such informationill not be misused but used to benefit both partners. The use

f new information for mutual benefits becomes easier when therust between partners is high. The cooperative behavior facilitatedy higher trust enables firms to survive greater stress and displaydaptability under changing conditions (Doz, 1996; Lorenz, 1988).igher trust between partners lowers transactions risks and infor-ation abuse online by partners and increases the chances that

ontractual terms will be honored online, and hence is likely tonhance the relationship between information sharing and flexi-ilities facilitated through e-procurement.

Based upon these arguments, we propose:

4. The effect of information sharing on purchase volume and mixexibilities will be stronger when the trust between the exchangeartners is high. That is, trust between exchange partners posi-ively moderates the relationship between information sharing andurchase volume and mix flexibilities.

.3. Performance implications of purchase volume and mixexibilities in e-procurement

Koste and Malhotra (1999) conducted a comprehensive reviewf antecedents and consequences of flexibility in the literature andalled for future studies to examine the relationship between flex-bility and other performance outcomes. Although prior researchas documented the impact of flexibility on firm performance andompetitiveness (Reichhart and Holweg, 2007; Anand and Ward,004; De Groote, 1994; Gerwin, 1993; Jordan and Graves, 1995;pton, 1997), we provide a unique perspective on how procure-ent flexibilities can impact performance at the process level. We

ontend that flexibility in business processes is a means to an end,nd in fact, may influence other dimensions of performance sig-ificantly. In other words, flexibility may be a predictor of otherimensions of performance, such as cost, quality, and delivery, thehree main drivers of performance in the supply-chain literatureThomas and Griffin, 1996). There are several reasons that mightxplain why flexibility might lead to better outcomes of cost, qual-ty, and delivery.

The first rationale is that material flow strongly impacts busi-ess performance, and organizational purchase volume and mixexibilities may enable timely modification of the volume and mixf products sourced from suppliers, and thereby can have a signif-cant impact on such flows. This is consistent with the precepts ofhe theory of swift and even flow (TSEF), which suggest that swiftow of materials will result in improved performance in terms ofost and quality (Schmenner and Swink, 1998).

The second reason is that flexibility in the purchasing processllows firms to be able to be responsive to changing customer needsnd requirements. In today’s marketplace marked with uncer-ainty in customer demands, firms that are able to better relayhe customer-induced changes to the supplier side (via flexibleurchasing agreements) will perform better on quality, costs, and

esponsiveness. Also, the ability to adjust the volume and mixf input goods dynamically using e-procurement enables compa-ies to lower transaction costs and employ competitive sourcingpportunities. Such efforts, typically, result in cost reductions of

anagement 30 (2012) 509–520 513

65% compared to “traditional” procurement transactions (Croom,2000).

Finally, one of the major benefits of flexibility in business pro-cesses is the responsiveness it affords to the members of thevalue-chain (Fisher, 1997). Such flexibility allows companies to usesubstitutable input goods and dynamically allocate different vol-ume and product mixes among suppliers, taking into considerationtheir competence and prior performance as well as the demand offinal products from the end-consumers.

Based on the above arguments, flexibility in volume and mix ishypothesized to have a beneficial impact on process performanceassessed along the dimensions of cost, quality, and responsiveness.We hypothesize:

H5 (a, b, c). Purchase volume and mix flexibilities facilitated bye-procurement will be positively associated with (a) lower costs,(b) higher quality, and (c) better responsiveness and delivery per-formance.

3. Sample and measures

We describe the sample, measures, and the psychometric prop-erties of the scales used in this section.

3.1. Sample

Our sample respondents consisted of procurement managersfrom the aerospace, automotive, electronics, and consumer prod-uct manufacturing industries in US. We solicited their responsesprimarily through mail and based on membership directories andmailing lists which included purchasing and procurement man-agers from Midwest US. Out of 400 individuals contacted, 141participated in the study yielding a response rate of 35.3%. Of these,only 131 responded to the complete survey, giving us a usable oreffective response rate of 32.8%. We informed the respondents thata benchmarking report will be made available to them at the con-clusion of the study. Participants were first introduced to the studyrequirements and then directed to the survey questionnaire in themail or directed to a website created for the purpose. There wasno statistical difference among the constructs used in this studybetween the responses obtained through the two methods. Theitems used in the survey are listed in Appendix A. Mailing wasdone in two waves and we conducted a standard two-sample Ttest examining the differences between the early and late waves asa way of testing non-respondent bias. We did not observe signif-icant differences between the two waves of data collection at the0.05 level of significance.

3.2. Construct validation

The measures employed in this study are adapted from extantliterature. Table 1 presents the definitions of the constructs alongwith their references. Appendix A provides the scale items, and reli-abilities of the constructs used in this study. Appendix B presentsthe descriptive statistics and correlation matrix. We used AMOS 5.0to implement a confirmatory factor analysis (CFA) described laterin this section.

3.2.1. ReliabilityReliability is an indication of measurement accuracy, which is

the extent to which instrumentation produces consistent or error-free results. In our study, all the developed scales demonstrated

reliabilities above the recommended threshold of 0.7 for cronbach’salpha (Nunnally and Bernstein, 1994). Appendix A illustrates reli-abilities, factor loadings, and the percent of variation extracted forall the constructs. We first conducted a factor analysis to get a

514 S. Devaraj et al. / Journal of Operations Management 30 (2012) 509–520

Table 1Definitions, constructs, and references.

Definitions References

ConstructSupplier customization This entails suppliers conforming to the needs of buyers and fulfilling them Doney and Cannon (1997)Information sharing Involves the extent to which suppliers and customers share private information with each

otherDoney and Cannon (1997)

Frequency The percent of transactions done onlineTrust Trusting suppliers online Doney and Cannon (1997)Purchase volume flexibility The ability of buyers to change the volume of products purchased with little penalty in time,

effort, cost, or performanceDuclos et al. (2003)Devaraj et al. (2004)

Purchase mix flexibility The ability of buyers to change the mix of products purchased with little penalty in time,effort, cost, or performanceBoth volume and mix flexibility enable firms to dynamically purchase products in response tochanges in market demands during the short life cycle of the product

Performance Performance comprises of relative cost, delivery, and qualityCost The extent to which online purchases provide cost advantages relative to traditional purchases Doney and Cannon (1997)

faster

form

pcrc

3

ridpa“armwqufpv

3

BBcfiTccgB0

3

iecftrcAc

Delivery The extent to which online purchases provide

traditional purchasesQuality The extent to which online purchases meet con

reliminary check for convergent and discriminant validities. Asan be seen from the table in Appendix C, all items loaded onto theirespective factors with high loadings and there were no significantross-loadings.

.2.2. Content validityContent validity is the degree to which items in an instrument

eflect the content generalization (Cronbach, 1951). As indicatedn Table 1, the items used in this survey were extracted from vali-ated scales in the extant literature. Furthermore, all the items wereresented in a pilot study to purchasing and procurement man-gers and validated by checking the content for their managerialreal-world” meaning. Feedback was sought from three researchersctively involved in e-procurement and supply-chain managementesearch. Next, we shared the questionnaire with three procure-ent managers. In both situations, we had a detailed conversationith the individuals, who made an assessment of the clarity of our

uestions. In some instances, the questions were modified basedpon their comments. The review of literature as well as the resultsrom the pilot study analyzed by a group of experts in the fieldrovided further reassurance about the adequacy of the contentalidity of our instrument (Boudreau et al., 2001).

.2.3. Convergent validityThe convergent validity of a scale can be checked using the

entler–Bonett coefficient (�) (Bentler and Bonnet, 1980). Theentler–Bonnet coefficient is the ratio of the difference between thehi-squared value of the null measurement model and the speci-ed measurement model to the chi-squared value of the null model.he null model has no hypothesized factor loading on a commononstruct in the confirmatory factor analysis. The Bentler–Bonnetoefficient value between 0.80 and 0.90 is acceptable and a valuereater than 0.90 indicates strong convergent validity (Bentler andonnet, 1980). The Bentler–Bonnet coefficient for our model is.916, which indicates strong convergent validity.

.2.4. Discriminant validityDiscriminant validity explores the extent to which a concept and

ts indicators differ from another concept and its indicators (Bagozzit al., 1991). This is achieved when measures of each constructonverge on their corresponding true scores (which are uniquerom other constructs) and can be tested by comparing the correla-ions between the pairs of dimensions (Venkatraman, 1989). This

equires a comparison of two CFA models, one with the correlationonstrained to equal one and another with an unconstrained model.

significantly lower �2 value for the model with the unconstrainedorrelation, when compared with the constrained model, provides

and more reliable delivery relative to Doney and Cannon (1997)

ance quality requirements Mentzer et al. (2001)

support for discriminant validity (Venkatraman, 1989). A �2 differ-ence value with an associated p-value less than 0.05 (Joreskog andSorbom, 1989) supports the discriminant validity criterion. Thisprocedure is repeated for all pairs of scales in the instrument (notpresented in the interests of brevity). The significances of all dif-ferences are less than 0.05, which indicates discriminant validity ofthe constructs employed in our study.

3.2.5. Overall model fitFinally, we conducted a CFA. Results of the CFA yielded a

goodness-of-fit index (GFI) of 0.928, adjusted goodness-of-fit(AGFI) of 0.801, Bentler–Bonnet delta of 0.916, root mean squareerror of approximation (RMSEA) of 0.04, and a chi-square to degreeof freedom ratio of 1.83. A GFI over the cut-off 0.9 and the AGFI overthe cut-off 0.8 imply good absolute fit. A RMSEA value means a lowresidual variance and hence a good parsimonious fit. RMSEA is closeto the cut-off value of 0.06 as suggested by Browne and Cudeck(2003). Thus, the CFA points to evidence of a good measurementmodel.

3.2.6. Common method varianceCommon method variance can be a potential source of bias in

survey research. We examined this issue in several ways. First, con-sistent with prior studies, we conducted Harman’s one-factor test(Podsakoff and Organ, 1986). A factor analysis performed on thevariables did not yield a single-factor solution. Hence, based onthis test the threat of common method bias is not significant. Thesecond test of common method variance was based on partial cor-relations (Podsakoff et al., 2003). We computed the first factor froma principal factor analysis of all the constructs and included this asa control variable in the analysis, since this factor is assumed tobe a good approximation of the common method bias. The addi-tion of the factor score as a control variable did not significantlychange the variation explained in the performance variables sug-gesting that common method variance is not a serious concern inour study (Podsakoff et al., 2003). In summary, based on tests ofreliability, validity, and overall model fit, there is strong supportfor the suitability of the constructs employed in this study.

4. Analyses and results

We employed structural equation modeling (SEM) to estimate

the proposed research model. To model the interactions betweensupplier customization and frequency of transactions as well asbetween information sharing and trust, we employed the crossproduct of the items of the separate constructs as the latent variable

S. Devaraj et al. / Journal of Operations M

Table 2Validity statistics.

Goodness of fit indices

Absolute fitChi-square/DF 1.34Goodness of Fit Index (GFI) 0.96Adjusted Goodness of Fit Index (AGFI) 0.84

Incremental fitNormed Fit Index (NFI) 0.91Incremental Fit Index (IFI) 0.91Non-Normed Fit Index (NNFI) 0.92Comparative Fit Index (CFI) 0.92

cfiopmrga

FomorIm(imtRsg

4

TSw

otsfl

TS

Parsimonious fitRoot Mean Square Error of Approximation (RMSEA) 0.05

apturing the interaction. Table 2 provides several goodness-of-t statistics to assess how well the specified model explains thebserved data from three aspects: absolute fit, incremental fit, andarsimonious fit (Maruyama, 1998, Tanaka, 1993). The absolute fiteasures of GFI, AGFI, and the Normed chi-square, which is the

atio of the chi-square divided by degrees of freedom indicate aood fit of the model to the data (Bentler and Bonnet, 1980; Bagozzind Yi, 1988).

Four incremental fit indices are also listed in Table 2. (i) Normedit Index (NFI), represents the proportion of total covariance amongbserved variables explained by a target model using a baseline nullodel, (ii) Incremental Fit Index (IFI) is the ratio of the discrepancy

f the proposed and baseline models over the difference of theirespective degrees of freedom (Bollen, 1989), (iii) Non-Normed Fitndex (NNFI), also known as Tucker–Lewis Index (TLI), examines

oment structures (Bentler and Bonnet, 1980; Bollen, 1989), andiv) Comparative Fit Index is one of the indices that determinesncremental fits and estimates non-centrality parameters of the

odel (Gefen et al., 2000). The values shown in the table suggesthat our proposed model fits the data well. Additionally, a smallMSEA value means a low residual variance and hence a good par-imonious fit and RMSEA. These values provide strong evidence ofood model fit.

.1. SEM results

A summary of the hypotheses test results is presented inable 3. The results of our analysis provide support for H1 (p < 0.01).upplier customization has a positive and significant relationshipith both purchase volume and mix flexibility.

We find partial support for the moderating effects of frequency

f transactions on the relationship between supplier customiza-ion and purchase volume and mix flexibilities. The relationship istatistically significant for purchase volume flexibility but not mixexibility. This finding suggests that while supplier customization

able 3ummary of the structural model.

Hypothesis

H1: Supplier customization → volume flexibility

Supplier customization → purchase mix flexibility

H2: Interaction of supplier customization and frequency → purchase volume flexibilityInteraction of supplier customization and frequency → purchase mix flexibility

H3: Information sharing → purchase volume flexibility

Information sharing → purchase mix flexibility

H4: Interaction of information sharing and trust → purchase volume flexibility

Interaction of information sharing and Trust → Purchase mix flexibilityH5A: Purchase volume flexibility → cost

Purchase mix flexibility → cost

H5B: Purchase volume flexibility → qualityPurchase mix flexibility → quality

H5 C: Purchase volume flexibility → deliveryPurchase mix flexibility → delivery

anagement 30 (2012) 509–520 515

in itself might have a beneficial impact on purchase volume flex-ibility, its effect is amplified when there is increased frequency ofonline transactions between supply chain partners. Thus, H2 is onlypartially supported.

We find statistical support for H3 (p < 0.01) confirming thatinformation sharing has a positive association with both purchasevolume and mix flexibility.

We also found evidence of positive moderating effects of truston the relationship between information sharing and both purchasevolume and mix flexibilities. In other words, when investments ininformation sharing are supported with trust between the part-ners, we see evidence of greater purchase volume flexibility andpurchase mix flexibilities. Thus, H4 is statistically supported.

Finally, H5 (a, b, and c) are all supported (p < 0.01), demon-strating that purchase volume and mix flexibilities facilitatedby e-procurement enable procurement cost, delivery and qualityimprovements in comparison to traditional modes of procurement.Finally, the control for size of the organization was statisticallysignificant with quality performance.

4.2. Alternate model testing: mediating role of purchase volumeand mix flexibilities

Fig. 2 illustrates the comparison of alternative models. We testedfor the mediating effect of purchase volume flexibility and pur-chase mix flexibility by implementing Baron and Kenny’s (1986)procedure to test for mediation.

Step 1: Use e-Procurement Performance as the dependent vari-able and Customization/Information Sharing as the independentvariable, test the statistical significance of this relationship.Step 2: Use Mix/Volume flexibility flow as the dependent vari-able and Customization/Information Sharing as the independentvariable to test for the statistical significance of this relationship.Step 3: Use e-Procurement Performance as the dependent variableand Customization/Information Sharing as independent variablesto test for the statistical significance of this relationship.Step 4: Test the effect of Customization/Information Sharing (inde-pendent variable) on e-Procurement Performance (dependentvariable) after controlling for Mix/Volume flexibility (indepen-dent variable). For Mix/Volume flexibility to completely mediatethe relationship between Customization/Information Sharingand e-Procurement Performance, the coefficient for Customiza-tion/Information Sharing should be statistically non-significant.

If all four steps are significant then the data are consistent withcomplete mediation. If steps 1–3 hold and step 4 does not, then thedata are consistent with partial mediation.

Expected relationship Path coefficient Supported?

+ 0.26 Yes; p < 0.01+ 0.17 Yes; p < 0.05+ 0.16 Yes; p < 0.05+ 0.07 No; p > 0.1+ 0.48 Yes; p < 0.01+ 0.31 Yes; p < 0.01+ 0.34 Yes; p < 0.01+ 0.14 Yes; p < 0.05+ 0.33 Yes; p < 0.01+ 0.24 Yes; p < 0.01+ 0.30 Yes; p < 0.01+ 0.22 Yes; p < 0.01+ 0.26 Yes; p < 0.01+ 0.11 Yes; p < 0.05

516 S. Devaraj et al. / Journal of Operations Management 30 (2012) 509–520

Fig. 2. Comparison of al

Table 4Model comparisons.

Hypothesized mediated model Direct effects model

(�2/DF) 1.34 2.77

mtpvMttp

tttcwtt

GFI 0.962 0.884AGFI 0.842 0.741

Consistent with these steps we find evidence for completeediation. In other words, information sharing and supplier cus-

omization provide no direct improvement on e-procurementerformance; their impacts are felt completely through purchaseolume and mix flexibilities. In testing structural equation models,arsh (1994) suggests comparing the research model to alterna-

ive models. Thus, we also tested a direct effects model to showhe direct effects of e-procurement process characteristics on e-rocurement performance.

A comparison of the two models is presented in Table 4.Based on the comparative model fit criteria, we observe that

he hypothesized mediating model fits the data significantly bet-er than the other model. Furthermore, the Normed chi-square (i.e.he (�2/DF) for our hypothesized mediating model is less than the

ut-off value of 2 as suggested by Byrne (1989). These results, alongith better goodness-of-fit indices for the mediated model, lead us

o believe that purchase volume and mix flexibilities play an impor-ant mediating role in explaining e-procurement performance.

ternative models.

4.3. A general path analytic model combining mediation andmoderation

There are many studies that combine moderation and mediationwithin the same research model. Normally, these are presentedas moderated mediation or mediated moderation. Recent workby Edwards and Lambert (2007) suggest that these analyticalapproaches might have important shortcomings that conceal thetrue form of the moderated and the mediated effects being studied.Their approach involves combining moderated regression analy-sis and path analysis to obtain better estimates of the moderatingand mediating coefficients. The general approach can be describedbelow:

1. Mediation is modeled as a path model with relationships amongvariables in the model captured as regression equations.

2. Moderation is modeled by supplementing the above regressionequations with the moderator variable and its product with theindependent variable and the mediator variable.

3. By using reduced form equations, obtained by substituting theregression equation for the mediator variable into the equationfor the dependent variable, direct, indirect, and total effects canbe computed.

4. The reduced form equations contain products of regressioncoefficients; therefore ordinary OLS cannot be used. Constrainednonlinear regression (CNLR) is used to generate coefficients frombootstrap samples.

tions M

oac(ararr

pv

5

mfiwaec

treWpptsdcefli

mawpaqwttiov

tsTitntosccip

S. Devaraj et al. / Journal of Opera

We used IBM SPSS 19 to conduct the above analysis. Resultsf this analysis showed an average coefficient for the moder-ting effect of frequency on the relationship between supplierustomization and volume flexibility and mix flexibility was 0.24significant) and 0.08 (not significant) respectively. Further, theverage coefficient for the moderating effect of trust on theelationship between information sharing and volume flexibilitynd mix flexibility was 0.32 (significant) and 0.16 (significant)espectively. These results affirm those obtained through SEM andeported in Table 3.

Taken together, the statistical support for all the hypothesesresented in this study point to the important role that purchaseolume and mix flexibilities play in the context of e-procurement.

. Implications, limitations, and future research

There are several academic and managerial implications stem-ing from the findings of this study. From a theoretical standpoint,

rst, there has been a need to examine the mechanisms throughhich business value of e-procurement is derived. We proposed

nd showed that using transaction cost economics and socialxchange theory, two mechanisms that drive flexibility are supplierustomization and information sharing.

Second, while purchase volume and mix flexibility is criticalo supply chain performance, it is the least studied in academicesearch and points to a gap in the literature. In today’s dynamicnvironment, the concept of flexibility is especially important.e define, and operationalize purchase volume flexibility and

urchase mix flexibility derived from e-procurement. Third, weropose and empirically test a nuanced perspective of the con-exts under which the two causal pathways from informationharing and supplier customization lead to flexibility. We find evi-ence for moderating effects of frequency and trust, two importantonstructs steeped in the transaction cost economics and socialxchange theory literature. While these two constructs have beenound significant for supply chain management, their role has beenargely unexplored in the context of purchase volume and mix flex-bilities.

Finally, there is still a need for academic research on the perfor-ance impacts of e-procurement. We show that purchase volume

nd mix flexibilities serve the important role of mediators throughhich eventual effects are realized on cost, quality, and deliveryerformance. Thus, flexibility in e-procurement can be viewed nots an end in itself, but rather a means to the end objectives of cost,uality, and responsiveness or delivery performance. In summary,hile extant literature contends that we still lack understanding of

he coordinated capabilities of an inter-organizational infrastruc-ure that will enable flexible supply chains and theory to guidets development (Gosain et al., 2005; Weill and Broadbent, 1998),ur study responds to this mandate by illustrating how purchaseolume and mix flexibilities impact e-procurement performance.

The only result that was not significant in our analysis washe hypothesized moderating role of frequency on the relation-hip between supplier customization and purchase mix flexibility.his result indicates that once a supplier has made the investmentn tools, assets and modified procedures, the recency of informa-ion made possible through higher frequency of transaction mayot provide any additional benefits as the a priori customiza-ion enables firms to modify the mix of products easily throughnline purchasing. From a managerial perspective, the relation-hips between the e-procurement characteristics and performance

an provide useful insights to procurement managers, supplyhain specialists, and practitioners. First we offer directions on thedentification and measurement of key characteristics that affecturchase volume and mix flexibility. Further, we demonstrate

anagement 30 (2012) 509–520 517

the business impact of enhancing purchase volume flexibility andpurchase mix flexibility. While there has been considerable atten-tion in practitioner publications on the value of these flexibilities,we assess the magnitude of its effect on measures of cost, qual-ity, and delivery by examining the coefficients in the estimationmodel. Thus, managers can decide if an investment in increasingpurchase volume and mix flexibilities is worthwhile in their con-texts. Furthermore, we demonstrate that purchase volume and mixflexibilities are instrumental for obtaining value in other procure-ment process dimensions, and thus managers must spend sufficientresources to facilitate them from a coordination perspective.

While there has been much discussion about agility in prac-titioner publications, there has been little empirical evidence ofthe business value of agility. To the extent that one of the keymanifestations of agility is flexibility, we demonstrate the businessvalue of agility by documenting the effect of flexibility on businessperformance such as cost, quality, and delivery. Based on infor-mation sharing and relying on each other’s strengths, electronicmanufacturing companies build relationships with their contractmanufacturers and this leads to strong and agile purchasing strat-egy (Mason et al., 2002). Consequently, collaborative systems thatfoster information sharing between buyer and suppliers providelarge companies with gains in flexibility and the ability to alterpurchasing capacity levels (Humphreys et al., 2001).

Finally, our study helps practitioners in understanding theimpact of the levers driving purchase volume and mix flexibilities,but more importantly, the contexts under which these levers workbest. That is, supply chain managers who find their organizationstransacting on a significant basis (frequency) with their customerswould do well to invest in customization, whereas their investmentin information sharing is more suitable when there is a high degreeof trust in the supplier–customer relationship.

The study has a few limitations as well which present opportuni-ties for future research in this area. First, respondents to our surveywere primarily from the aerospace, automotive, electronics, andconsumer products manufacturing industries. As such, our resultsare generalizable to such a population and further research mightbe warranted to demonstrate the efficacy of our research modelin a wider cross-section. Second, it is conceivable that there aremany other dimensions of procurement performance than thosethat we included in the study such as strategic importance ofe-procurement to the firm’s overall strategy. To the extent thatindustry type and size of the firm might serve as proxies for strate-gic necessity of e-procurement, our analysis accounts for them inthe form of control variables. Our selection depicts a parsimoniouscollection of constructs based on extant literature as well as ourdiscussions with procurement managers. It is not our intent toexamine the various forms of e-procurement systems in practicebut address the more fundamental issue of what underlying char-acteristics affect performance and how they do it. Future studiesmight examine the various forms of procurement systems. This isagain an opportunity for future research to expand the domain ofthe relationships examined in this study. Finally, we have consid-ered the buyer side perspective only in this empirical study. Furtherstudies might explore the buyer–seller dyad and its implications fore-procurement performance.

In conclusion, while e-procurement technologies have becomepervasive in business, there has been little academic inquiryto understand the process by which they confer performancegains. We proposed and empirically examined how two keymechanisms — supplier customization, and information shar-ing — impact purchase volume and mix flexibilities, which in

turn affects e-procurement performance. We hope our studyspurs further interest and research to examine other path-ways through which e-procurement technologies provide businessvalue.

5

A

N

A

18 S. Devaraj et al. / Journal of Operations Management 30 (2012) 509–520

ppendix A.

Scale items and reliabilities of constructsFactor loading

Supplier customization (adapted from Doney and Cannon, 1997) ( ̨ = 0.742, variation explained = 66.42%)CUST1 0.690 Just for us my online supplier is willing to customize its productsCUST2 0.858 Just for us my online supplier is willing to change its processCUST3 0.884 Just for us my online supplier is willing to customize services, tools and equipment

Information sharing (adapted from Doney and Cannon, 1997) ( ̨ = 0.736, variation explained = 65.84%)INFO1 0.852 Our online suppliers share proprietary information with our firmINFO2 0.794 Our online suppliers will share confidential information to help usINFO3 0.787 Our online suppliers make online catalog information available

Trust (adapted from Doney and Cannon, 1997) ( ̨ = 0.936, variation explained = 73.95%)TRST1 0.704 Our online suppliers keep promises it makes to our firmTRST2 0.915 Our online suppliers are not always honest with us (reverse coded)TRST3 0.961 We believe the information that our online suppliers provide us.TRST4 0.701 Our online suppliers are genuinely concerned that our business succeedsTRST5 0.961 When making important decisions, our online suppliers consider our welfare as well as its ownTRST6 0.915 We trust that our online suppliers keep our best interests in mindTRST7 0.783 Our online suppliers are trustworthyTRST8 0.892 We find it necessary to be cautious with online suppliers (reverse coded)

Cost (adapted from Doney and Cannon, 1997) ( ̨ = 0.907, variation explained = 84.30%)COST1 0.949 I save money when I purchase online compared to traditional purchasesCOST 2 0.920 Online purchasing is costlier than traditional purchasing (reverse coded)COST 3 0.884 Our online cost per transaction relative traditional purchasing has been reduced

Delivery (adapted from Doney and Cannon, 1997) ( ̨ = 0.725, variation explained = 72.69%)DELV1 0.853 By ordering online, delivery is faster than traditional purchasingDELV2 0.853 By ordering online, delivery is more reliable than traditional purchasing

Quality (adapted from Mentzer et al., 1997; Parasuraman et al., 1988) ( ̨ = 0.707, variation explained = 72.69%)QUAL1 0.879 Products ordered online meet technical requirements compared to traditional purchasingQUAL2 0.607 Products ordered online are rarely nonconforming compared to traditional purchasing.QUAL3 0.800 I believe that what I ask for is what I get in online purchasing compared to traditional purchasing

Purchase volume flexibility and purchase mix flexibility (adapted from Devaraj et al., 2004) ( ̨ = 0.723, 0.708) variation explained = 62.46% and 60.05%)PVF11 Online purchasing increases our purchase volume flexibility relative to traditional purchasingPVF12 Online services or suppliers provide me with flexibility in purchasingPMF13 Online purchasing increases our purchase mix flexibility relative to traditional purchasingPMF14 Online purchasing provides me with flexibility in purchasing

ote: A seven point Likert scale from 1 to 7 (strongly disagree to strongly agree) was used for all study constructs.

ppendix B.

Correlation matrix and descriptive statistics for constructs

Construct Mean SD (1) (2) (3) (4) (5) (6) (7) (8) (9)

(1) Customization 6.48 0.56(2) Information sharing 6.28 0.625 0.559**

(3) Frequency 16.71 6.41 −0.176* −0.480(4) Trust 6.20 0.43 0.108 0.023 −0.021(5) Purchase volume flexibility 6.47 0.48 0.318** 0.410** −0.158 0.081(6) Purchase mix flexibility 6.32 0.42 0.187* 0.318** −0.078 −0.098 0.259**

(7) Cost 6.38 0.57 0.323** 0.457** −0.176* 0.047 0.322** 0.246**

(8) Quality 6.46 0.41 0.216* 0.408** −0.277* 0.001 0.304** 0.294** 0.329**

(9) Delivery 6.18 0.52 0.352** 0.345** −0.220* 0.217* 0.393** 0.180* 0.108 0.180*

(10) Size 54938.09 48647.15 0.116 0.237 0.171 −0.148 0.164 0.011 0.100 0.228 0.122* Correlation is significant at the 0.05 level (2-tailed).

** Correlation is significant at the 0.01 level.

A

F

E

R

A

A

B

B

B

B

B

B

B

B

B

C

C

C

C

C

D

D

D

S. Devaraj et al. / Journal of Operations Management 30 (2012) 509–520 519

ppendix C.

actor analyses showing unidimensionality, and convergent and discriminant validityComponent

1 2 3 4 5 6 7 8

TRST1 0.652 −0.054 −0.085 −0.161 −0.064 0.115 −0.050 0.087TRST2 0.937 0.115 0.036 0.221 −0.040 0.072 0.035 0.127TRST3 0.942 0.006 0.040 0.103 −0.033 0.062 0.096 0.084TRST4 0.673 −0.096 −0.126 −0.111 −0.031 0.087 −0.062 −0.021TRST5 0.952 0.006 0.040 0.103 −0.033 0.062 0.096 0.073TRST6 0.883 −0.034 −0.034 −0.145 −0.040 0.067 −0.008 −0.023TRST7 0.742 0.019 0.076 0.028 0.161 0.092 −0.139 −0.049TRST8 0.883 0.134 0.078 0.110 −0.047 0.086 0.075 0.061COST1 0.038 0.903 0.022 0.093 0.110 0.039 0.152 0.143COST2 0.092 0.849 0.187 0.135 0.149 0.092 0.141 0.168COST3 −0.038 0.863 0.122 0.079 0.031 0.015 −0.049 −0.069QUAL1 0.086 −0.016 0.901 −0.049 −0.034 0.202 0.042 0.073QUAL2 −0.006 0.104 0.875 0.025 0.178 −0.059 −0.081 −0.063QUAL3 −0.098 0.111 0.563 0.209 0.220 −0.020 0.179 0.038CUST1 0.161 0.256 0.178 0.551 0.091 0.142 −0.027 0.051CUST2 0.123 0.027 −0.046 0.684 0.135 0.198 0.078 0.083CUST3 0.086 0.112 −0.078 0.623 0.153 0.194 0.029 0.034INFO1 0.092 0.202 0.121 0.158 0.641 0.094 −0.104 −0.126INFO2 0.040 0.273 0.101 0.152 0.467 0.190 0.040 0.069INFO3 −0.096 0.045 0.050 0.129 0.848 0.133 0.165 0.153DELV1 0.134 −0.047 0.108 0.084 0.172 0.826 0.143 0.182DELV2 0.134 −0.006 0.073 0.140 0.063 0.787 0.126 0.137FLEX1 0.116 0.281 0.193 0.135 0.041 0.161 0.727 0.349FLEX2 0.053 0.123 0.065 0.034 0.137 0.089 0.902 0.391FLEX3 0.055 0.177 0.033 0.138 0.062 0.121 0.341 0.734FLEX4 0.072 0.125 0.074 0.092 0.114 0.142 0.186 0.631

xtraction method: Principal Component Analysis. Rotation method: Varimax with Kaiser Normalization.

eferences

nderson, J., Narus, J., 1990. A model of distributor firm and manufacturer firmworking relationships. Journal of Marketing 54 (1), 42–58.

nand, G., Ward, P.T., 2004. Fit, flexibility and performance in manufacturing: copingwith dynamic environments. Production and Operations Management 13 (4),369–385.

agozzi, R.P., Yi, Y., 1988. On the evaluation of structural equation models. Academyof Marketing Science 16 (1), 74–94.

agozzi, R.P., Yi, Y., Phillips, L.W., 1991. Assessing construct validity in organizedresearch. Administrative Science Quarterly 36 (3), 421–458.

aron, R.M., Kenny, D.A., 1986. The moderator–mediator variable distinction insocial psychological research: conceptual, strategic and statistical consider-ations. Journal of Personality and Social Psychology 51, 1173–1182.

entler, P.M., Bonnet, D.G., 1980. Significance tests and goodness of fit in the analysisof covariant structures. Psychological Bulletin 88 (3), 588–606.

ollen, K.A., 1989. A new increment fir index for general structural equation models.Sociological Methods and Research 17, 303–316.

oudreau, M., Gefen, D., Straub, D.W., 2001. Validation in IS research: a state-of-the-art assessment. MIS Quarterly 25 (1), 1–24.

radley, A., 2005. E-procurement: a long way to go. Supply Management 10 (20),15.

rowne, M., Cudeck, R., 2003. Alternative ways of assessing model fit. In: Bollen, K.A.,Long, J.S. (Eds.), Testing Structural Equation Models. Sage Publications, London,U.K.

yrne, B.M., 1989. A Primer of LISREL: Basic Applications and Programming forConfirmatory Factor Analytic Models. Springer-Verlag, New York, NY.

achon, G., Fisher, M., 2000. Supply chain inventory management and the value ofshared information. Management Science 46 (8), 1032–1048.

ao, Q., Dowlatshahi, S., 2005. The impact of alignment between virtual enterpriseand information technology on business performance in an agile manufacturingenvironment. Journal of Operations Management 23, 531–550.

hen, I.J., Paulraj, A., Lado, A.A., 2004. Strategic purchasing, supply management andfirm performance. Journal of Operations Management 22 (5), 505–523.

ronbach, L.J., 1951. Coefficient alpha and the internal structure of tests. Psychome-trika 16, 297–334.

room, S., 2000. The impact of Web-based procurement on the management ofoperating resources supply. The Journal of Supply Chain Management 36 (1),4–13.

e Groote, X., 1994. The flexibility of production processes: a general framework.Management Science 40 (7), 933–945.

Doney, P.M., Cannon, J.P., 1997. An examination of the nature of trust in buyer–sellerrelationships. Journal of Marketing 61, 35–51.

Doz, Y.L., 1996. The evolution of cooperation in strategic alliances: initial conditionsor learning processes. Strategic Management Journal 17, 55–83.

Duclos, L.K., Vokurka, R.J., Lummus, R.R., 2003. A conceptual model of supply chainflexibility. Industrial Management & Data Systems 103 (6), 446–457.

Edwards, J.R., Lambert, L.S., 2007. Methods for integrating moderation andmediation: a general analytical framework using moderated path analysis. Psy-chological Methods 12 (1), 1–22.

Fisher, M.L., 1997. What is the right supply chain for your product? Harvard BusinessReview 75 (2), 105–116.

Gefen, D., Straub, D.W., Boudreau, M.C., 2000. Structural equation modeling andregression: guidelines for research practice. Communications of AIS 4 (7), 71–78.

Gerwin, D., 1993. Manufacturing flexibility: a strategic perspective. ManagementScience 39 (4), 395–410.

Gosain, S., Malhotra, A., El Sawy, O.A., 2005. Coordinating for flexibility in e-businesssupply chains. Journal of Management information Systems 21 (3), 7–46.

García-Dastugue, S.J., Lambert, D.M., 2003. Internet-enabled coordination in thesupply chain. Industrial Marketing Management 32 (3), 251–263.

Handfield, R.B., Bechtel, C., 2002. The role of trust and relationship structure inimproving supply chain responsiveness. Industrial Marketing Management 31,367–382.

Hill, C.A., Scudder, G.D., 2002. The use of electronic data interchange for supplychain coordination in the food industry. Journal of Operations Management 20,375–387.

Hsieh, Y., Chiu, H., Hsu, Y., 2008. Supplier market orientation and accommodation ofthe customer in different relationship phases. Industrial Marketing Management37, 380–393.

Hsu, C., Kannan, V.R., Tan, K., Leong, G.K., 2008. Information sharing, buyer–supplierrelationships, and firm performance: a multi-region analysis. International Jour-nal of Physical Distribution & Logistics Management 38 (4), 296–310.

Humphreys, P.K., Shiu, W.K., Chan, F.T.S., 2001. Collaborative buyer–supplier rela-tionships in Hong Kong manufacturing firms. Supply Chain Management: AnInternational Journal 6 (4), 152–162.

Jack, E.P., Raturi, A., 2002. Sources of volume flexibility and their impact on perfor-mance. Journal of Operations Management 20 (5), 519–548.

Jap, S., Mohr, J.J., 2002. Leveraging Internet technologies in B2B relationships.California Management Review 44 (4), 24–38.

Johnson, P.F., Klassena, R.D., Leendersa, M.R., Awaysheha, A., 2007. Utilizing e-business technologies in supply chains: the impact of firm characteristics andteams. Journal of Operations Management 25 (6), 1255–1274.

Jordan, W.C., Graves, S.C., 1995. Principles on the benefits of manufacturing: process

evaraj, S., Wei, J., Krajewski, L., 2007. Impact of eBusiness technologies on opera-

tional performance: the role of production information integration in the supplychain. Journal of Operations Management 25 (6), 1199–1216.

evaraj, S., Hollingworth, D.G., Schroeder, R.G., 2004. Generic manufacturing strate-gies and plant performance. Journal of Operations Management 22 (3), 313–333.

flexibility. Management Science 41 (4), 577–594.Joreskog, K.G., Sorbom, D., 1989. LISREL 7: A Guide to the Program and Application.

SPSS Inc., Chicago.Kekre, S., Srinivasan, K., 1990. Broader product line: a necessity to achieve success?

Management Science 36 (10), 1216–1231.

5 tions

K

K

K

L

M

M

M

M

M

M

N

O

P

P

P

P

P

P

R

R

R

20 S. Devaraj et al. / Journal of Opera

im, K.K., Umanath, N.S., Kim, B.H., 2005. An assessment of electronic informationtransfer in B2B supply–channel relationships. Journal of Management Systems22 (3), 293–320.

ollock, P., 1994. The emergence of exchange structures: an experimental studyof uncertainty, commitment and trust. American Journal of Sociology 100,313–345.

oste, L.L., Malhotra, M.K., 1999. A theoretical framework for analyzing the dimen-sions of manufacturing flexibility. Journal of Operations Management 18 (1),75–93.

orenz, E., 1988. Neither friends nor strangers: informal networks of subcontractingin French Industry. In: Gambetta, D. (Ed.), Trust: Making or Breaking CooperativeRelations. Basil-Blackwell, New York, pp. 194–210.

arsh, H.W., 1994. Confirmatory factor analysis models of factorial invariance: amultifaceted approach. Structural Equation Modeling 1 (1), 5–34.

ason, S.J., Cole, M.H., Ulrey, B.T., Yan, L., 2002. Improving electronics manufactur-ing supply chain agility through outsourcing. International Journal of PhysicalDistribution & Logistics Management 32 (7), 610–620.

aruyama, G.M., 1998. Basics of Structural Equation Modeling. Sage Publishers,Thousand Oaks, CA.

entzer, J.T., Flint, D.J., Hult, G.T.M., 2001. Logistics service quality as a segment-customized process. Journal of Marketing 65 (4), 82–104.

entzer, J.T., Rutner, S.M., Matsuno, K., 1997. Application of the means-end valuehierarchy model of understanding logistics service quality. International Journalof Physical Distribution and Logistics Management 27 (9/10), 230–243.

uffato, M., Payaro, A., 2004. Implementation of e-procurement and e-fulfillmentprocesses: a comparison of cases in the motorcycle industry. International Jour-nal of Production Economics 89, 339–351.

unnally, J.C., Bernstein, I.H., 1994. Psychometric Theory, 3rd ed. McGraw-Hill, NewYork, NY.

lson, J.R., Boyer, K.K., 2003. Factors influencing the utilization of Internet purchas-ing in small organizations. Journal of Operations Management 21 (2), 225–245.

arasuraman, A., Zeithaml, V.A., Berry, L.L., 1988. SERVQUAL—A multiple-item scalefor measuring consumer perceptions of service quality. Journal of Retailing 64(1), 12–40.

arkhe, A., 1993. Strategic alliance structuring: a game theoretic and transactioncost examination of interfirm cooperation. Academy of Management Journal36, 794–829.

odsakoff, P.M., MacKenzie, S.B., Jeong-Yeon, L., Podsakoff, N.P., 2003. Commonmethod biases in behavioral research: a critical review of the literature andrecommended remedies. Journal of Applied Psychology 88 (5), 879–903.

odsakoff, P.M., Organ, D.W., 1986. Self reports in organizational research: problemsand prospects. Journal of Management 12 (4), 531–544.

oppo, L., Zenger, T., 1998. Testing alternative theories of the firm: transac-tion cost, knowledge-based, and measurement explanations for make-or-buy decisions in information services. Strategic Management Journal 19,853–877.

remkumar, G., Ramamurthy, K., 1995. The role of interorganizational and organi-zational factors on the decision mode for the adoption of interorganizationalsystems. Decision Sciences 26 (3), 303–336.

eichhart, A., Holweg, M., 2007. Creating the customer-responsive supply chain:a reconciliation of concepts. International Journal of Operations & ProductionManagement 11 (27), 1144–1172.

osenzweig, E.D., Roth, A.V., 2007. B2B seller competence: construct developmentand measurement using a supply chain strategy lens. Journal of OperationsManagement 25, 1311–1331.

osenzweig, E., Roth, A.V., Dean, J.W., 2003. The influence of an integration strategyon competitive capabilities and business performance: an exploratory study of

Management 30 (2012) 509–520

consumer products manufacturers. Journal of Operations Management 21 (4),437–456.

Sahin, F., Robinson, E.P., 2005. Information sharing and coordination in make-to-order supply chains. Journal of Operations Management 23, 579–598.

Sahin, F., Robinson, F.P., 2002. Flow coordination and information sharing in sup-ply chains: Review, implications, and directions for future research. DecisionSciences 33 (4), 505–536.

Sambamurthy, V., Bharadwaj, A., Grover, V., 2003. Shaping agility through digitaloptions: reconceptualizing the role of information technology in contemporaryfirms. MIS Quarterly 27 (2), 237–263.

Sanders, N., 2007. An empirical study of the impact of e-business technologies onorganizational collaboration and performance. Journal of Operations Manage-ment 25 (6), 1332–1347.

Salvador, F., Rungtusanatham, M., Forza, C., Trentin, A., 2007. Mix flexibilityand volume flexibility in a build-to-order environment: synergies and trade-offs. International Journal of Operations & Production Management 11 (27),1173–1191.

Schmenner, R.W., Swink, M.L., 1998. On theory in operations management. Journalof Operations Management 17 (1), 97–113.

Seabright, M., Levinthal, M., Fichman, M., 1992. Role of individual attachments inthe dissolution of interorganizational relationships. Academy of ManagementJournal 35, 122–160.

Sia, S.W., Kjoh, C., Tan, C.X., 2008. Strategic maneuvers for outsourcing flexibility:an empirical assessment. Decision Sciences 39 (3), 407–443.

Suarez, F.F., Cusumano, M.A., Fine, C.H., 1996. An empirical study of manufactur-ing flexibility in printed circuit board assembly. Operations Research 44 (1),223–240.

Subramani, M., 2004. How do suppliers benefit from information technology use insupply chain relationships? MIS Quarterly 28 (1), 45–73.

Swafford, P.M., Ghosh, S., Murthy, N., 2006. The antecedents of supply chain agility ofa firm: scale development and model testing. Journal of Operations Management24, 170–188.

Talluri, S., Sarkis, J., 2002. A model for performance monitoring of suppliers. Inter-national Journal of Production Research 40 (16), 4257–4269.

Tanaka, J.S., 1993. Multifaceted conceptions of fit in structural equation models. In:Bollen, K.A., Long, J.S. (Eds.), Testing Structural Equation Models. Sage Publishers,Newbury Park, CA.

Thomas, D.J., Griffin, P.M., 1996. Coordinated supply chain management. EuropeanJournal of Operations Research 94 (1), 1–15.

Teo, T.S.H., Yu, Y., 2005. Online buying behavior: a transaction cost economics per-spective. Omega 33, 451–465.

Upton, D.M., 1997. Process range in manufacturing: an empirical study of flexibility.Management Science 43 (8), 1079–1092.

Vaidyanathan, G., Devaraj, S., 2008. The role of quality in e-procurement per-formance: an empirical examination. Journal of Operations Management 26,407–425.

Venkatraman, N., 1989. Strategic orientation of business enterprises: the construct,dimensionality and measurement. Management Science 35 (8), 942–962.

Vickery, S., Dröge, C., Germain, R., 1999. The relationship between product cus-tomization and organizational structure. Journal of Operations Management 17(4), 377–391.

Weill, P., Broadbent, M., 1998. Leveraging the New Infrastructure: How Mar-

ket Leaders Capitalize on Information. Harvard Business School Press, Boston,MA.

Young-Ybarra, C., Wiersema, M., 1999. Strategic flexibility in information technol-ogy alliances: the influence of transaction cost economics and social exchangetheory. Organization Science 10 (4), 439–459.