An empirical study of EDI trading partner selection
criteria in customer-supplier relationships
Rebecca Angeles (Ph.D.)
a,*, Ravinder Nath (Ph.D.)
b,1aDepartment of Information and Decision Sciences, School of Business, Montclair State University Upper Montclair, NJ 07043, USA bDepartment Chair, Information Systems and Technology, Director, The Joe Ricketts Center, College of Business,
Creighton University 2500 California Plaza, Omaha, NE 68178, USA
Received 12 July 1999; accepted 26 September 1999
Abstract
Electronic data interchange (EDI)-enabled trading partnerships are even more important now that EDI and electronic commerce-based technologies are underlying long-term strategic business partnerships. This study investigates the trading partner selection criteria used by ®rms in a customer-supplier dyad and their relative importance according to EDI implementation level is also established. Using the survey method implementing paired questionnaires for a dyad of customer-supplier ®rms, the study gathered data from 152 respondent ®rms. Factor analysis yielded six factors in trading partner selection: strategic commitment, trading partner ¯exibility, joint partnering for EDI, readiness for high-level EDI, EDI infrastructure, and communications. MANOVA andt-tests were used to test differences in the means of the responses of customer and supplier ®rms to the selection criteria. Overall, customer ®rms assigned higher means to all six factors than did the supplier ®rms. The gap between the two groups of ®rms were widest for the factors readiness for high-level EDI, trading partner ¯exibility, and communications.#2000 Elsevier Science B.V. All rights reserved.
Keywords:Electronic data interchange (EDI); Internet-EDI; Inter-organizational systems; Trading partnerships; Strategic alliances; Customer ®rms; Supplier ®rms
1. Introduction
The environment in which electronic data inter-change (EDI) has been used has inter-changed: we now have a `digital nervous system' made possible by wired organizations and Internet-based technologies that allow ®rms to exchange information, act, and
react within shorter windows of time. Traditional EDI has evolved to Internet-EDI, which means using cheaper technology to allow small- and medium-sized businesses to participate in the electronic marketplace. A far more discriminating and demanding customer base that is now accustomed to highly individualized products and services is driving companies towards strategies enhancing mass customization.
Highly integrated supply chain management (SCM) and accompanying logistics services have now become the basis of competition in the increasingly electronic and web-driven marketplace. One de®nition of SCM is the integrative approach that covers the ¯ow
*Corresponding author. Tel.:1-973-655-5336; fax:1-973-655-7678.
E-mail addresses: angelesr@mail.montclair.edu (R. Angeles), rnath@creighton.edu (R. Nath)
1Tel.:1-402-280-2439; fax:1-402-280-5565.
of the channel from the time raw materials are sourced from the ®rst supplier up to the time the ®nished products reach the last customer and beyond, includ-ing the disposal process in some systems [14]. Firms that have used EDI are now reexamining themselves so that they could move into the higher trajectory of implementing more highly integrated systems with EDI components that provide robust electronic links throughout the supply chain. New organizational forms such as the `extended' or `agile' enterprise have been emerging to allow for tighter links among strategic partners±customers, suppliers, or other third party service providers±that decide to dovetail their capabilities to provide a seamless and electro-nically enabled closed loop of unimpeded business processes.
The question, then, arises for the ®rm: how many strategic trading partners should it partner with under more demanding market conditions? In looking at the customer-supplier dyad, Bakos and Brynjolfsson [3] found that ®rms are ®nding it more pro®table to work with a smaller number of suppliers. Use of information technology is lowering coordination costs (i.e. search costs, the cost of setting up a relationship, and trans-action costs), thus, inviting a situation where a ®rm could deal with a potentially larger number of sup-pliers. However, with the movement towards more integrated and agile `extended enterprises', hub ®rms (i.e. ®rms that initiate EDI linkages) have been forced to provide incentives to their suppliers to make non-contractible investments in information sharing, qual-ity initiatives, and innovation to enable them to ful®ll the requirements of more tightly connected and inte-grated information networks. It makes better sense, then, to reduce the number of suppliers one could deal with as `partners'.
While the rationale for selecting trading partners at the least transaction cost has been suggested by transaction cost economics [4], there has been no comprehensive framework presented addressing the organizational aspects of EDI trading partnerships that directly affect the selection process. Many ®rms doing EDI have had previous relationships with their trading partners even before electronic linkages were estab-lished between them. In the case of these ®rms, the EDI connection is merely an extension of this relationship. It is also understood that most EDI net-works involve the `cluster' pattern where the hub
company has sets of trading partners±customers and suppliers for particular line product lines. Their trad-ing partners, in turn, form their own network clusters with other customers or suppliers in the marketplace. A global view of these relationships will present the image of a series of interconnected electronically linked spider-like `webs'.
This study seeks to investigate the trading partner selection criteria used by ®rms in a customer-supplier dyad. The importance of these selection criteria is also investigated across the different levels of EDI imple-mentation.
EDI is the direct computer-to-computer exchange of information stored in standard formatted business documents, such as invoices, bills of lading, purchase orders, etc., among organizations participating in a trading partnership network [10,18,21,33,40,44]. According to Sprague and McNurlin [67], EDI systems are a speci®c form of `cooperative' sys-tems±automated systems shared by two or more organizations. These are also referred to as `Inter-organizational Information Systems (IOS)' [6,9,11, 12,46]. Certain characteristics distinguish EDI IOSs [43] from other systems:
1. The availability of a computer system that is compatible with that of one's trading partner is essential if a direct linkage between partners will be established and a third party service network such as a value-added network (VAN) is not used [16].
2. The adoption of data and communications stan-dards is critical in implementing IOSs.
3. Education is important. Trading partner firm personnel need to be informed about the require-ments and implications of such systems to ensure success.
4. The involvement of a third-party entity such as a value-added network or VAN is typical. They train IOS participants, maintain EDI standards, and connect trading partners.
5. Work activities must be synchronized among IOS participants, especially when structured data format standards need to be changed or updated. 6. Work processes need to be reevaluated and
possibly, reengineered.
share information. The initiator of the IOS is usually the powerful entity since it selects the other firms that will be allowed to join the network.
There are several reasons for carefully selecting one's EDI trading partners. First, it is well established that ®rms need to be capable of mounting projects involving highly integrated EDI. Examples of EDI requiring considerable integration include `Just-in-Time (JIT)' in manufacturing; `Quick Response' that adapts those principles in the apparel retailing industry and uses complementary technologies such as barcod-ing; and `Ef®cient Consumer Response' that also adapts the concepts but in the grocery and food distribution business, and in all these industries, the practice of vendor-management inventories [15,19]. These involve `channel partnerships' ideally cemen-ted by mutual trust and a shared sense of collaboration and common destiny [27]. A channel partnership occurs when `. . . the parties agree on objectives,
policies, and procedures for ordering and physical distribution of the supplier's products. . ..' [8].
Second, as more ®rms cluster into such technology-backed partnerships based on IOSs, the locus of market rivalry will shift from competition between ®rms to competition between IOSs [22]. It is impera-tive, therefore, for the IOS to effectively manage its logistical, ®nancial, technical, and design interdepen-dencies to allow it to behave as one effective unit. Doing so with carefully selected partners will simplify this complex coordination task. Studies have shown that EDI hub ®rms that make specialized investments such as in hardware, software, transmission facilities, data/databases, and expertise tend to improve the `transactional climate' or sentiments that exist between itself and its trading partners [34].
Third, the technological issues of managing even high-end EDI networks will eventually be inconse-quential and the basis for competitive advantage will be on `relationship management' [30]. Riggins and Mukhopadhyay [39] have promoted the concept of `business partner reengineering' to strengthen an IOS's ability to consolidate its resources, coordinate its actions, streamline its procedures, and forge increasing trust levels among partners. Subsequent research studies show that cooperation among EDI trading partners shown, for instance, by increasing the
amount of information exchanged, is signi®cantly related to channel performance [47].
Fourth, although collaborative IOSs is the vision behind the clustering of ®rms into networks, the reality in the marketplace is that of a `lopsided' IOS, with the power accruing to the hub of the network, which is usually a large customer with a prominent presence in industry. What EDI networks have engendered in these cases is con¯ict and disgruntled compliance with the mandate of the more powerful trading partner. In the apparel retailing industry, retailers enjoy reduced inventory carrying costs, but the suppliers have had to bear EDI and inventory costs. Increased sales for the suppliers usually compensate for the increased inventory carrying costs.
The economic and ®nancial aspects of maintaining an EDI network also argue for selectivity in evaluating potential trading partners. One study, using extensive analysis of the EDI network with a two-level hier-archical model consisting of one buyer and a number of heterogeneous, competing suppliers, arrived at the conclusion that it is better for a dominant buyer to subsidize its suppliers especially when the buyer stands to derive a signi®cant reduction in its operating expenses and when the suppliers' EDI adoption costs are fairly high [49]. This same study also found that a buyer's pro®t potential from production will increase as the number of suppliers hooking up the EDI net-work increases and that adoption by a very productive supplier can have a great impact on the buyer's pro®t-ability. The problem, though, is that the marginal pro®t increment for the participating suppliers decreases monotonically as more of their competitors are signed into the network.
The results of these studies should alert EDI coor-dinators and managers to allocate the costs of sub-sidization in the feasibility stage of the project and to tolerate suboptimal bene®ts from the network for a period of time. Because of this, it is imperative that an EDI network hub carefully choose trading partners. There is, however, one way for a hub to work±just say± `if you want us to sell your products, join our net-work'.
There is a paucity of research on just how hub ®rms choose their trading partners (TP). Prior history of doing business together, volume of business or sales generated, and possession of EDI experience have been the most common bases of the selection. The objectives of our study, therefore, are to:
1. Identify selection criteria that ®rms use in establishing EDI linkages;
2. Compare and contrast the selection criteria that customer and supplier firms use in creating EDI partnerships;
3. Compare and contrast the criticality of the selection criteria according to level of EDI implementation, and
4. Provide prescriptive guidelines for establishing EDI linkages.
1.1. Definition of `level of EDI implementation'
There are three EDI implementation levels: (a) Level 1: simple data exchange without integration into any of the ®rm's internal processes; (b) Level 2: data is being exchanged between applications of the two ®rms; and (c) Level 3: business process reengi-neering is performed to allow EDI to change the way the ®rm performs its activities [17]. A comprehensive study noting the number of ®rms at each level of implementation has not been conducted across all EDI users in the US. However, it has been observed that the higher the level of EDI implementation, the greater the bene®ts experienced by the network parti-cipants [7].
2. Methodology
Data for this research were collected using a ques-tionnaire instrument.
2.1. Questionnaire instrument for EDI trading partnerships
There is no comprehensive study as of yet in-dicating the usefulness of selecting trading partners for vertical EDI alliances. One indicative study, though, by Williams et al., [50] found that ®rms that undertook EDI trading partner veri®cation programs were more likely to achieve greater range (i.e. the percentage of trading partners with whom a ®rm shares EDI information), depth (i.e. percentage of data processing done through EDI in relation to manual processing), and width of EDI usage (i.e. extent to which a ®rm uses EDI for multiple business applications).
This study looks at electronic partnerships that involve the supply value-chain IOS as de®ned by Kumar and van Dissel [26], which support custo-mer-supplier relationships also referred to as pipeline
management systems [29]. These relationships
embody sequential interdependencies or interactions that may extend from tracking EDI-based purchase orders, searching the database for adjacent partners in the chain to forecast customer needs, to transferring computer-aided-design (CAD)-based speci®cations from customer to supplier. Firms that agree to parti-cipate in these vertical SCM alliances primarily seek to reduce uncertainties in the supply chain in order to gain cost, cycle time, and quality advantages over rival supply chains in the industry [32].
While many EDI trading partner relationships have been responses to customer directives to be EDI compliant [25,48], the transition from a competitive to a more cooperative posture in dealing with potential trading partners appears to be taking place [23,26]. This study is premised on the need to carefully select trading partners, particularly those involved in highly integrated alliances, to ensure the survival of the relationship over the long term.
Table 1
Selection criterion items and related literature
Criterion References
1. Trading partner's flexibility and Willingness to meet your information and other needs
Hendrick and Ellram, 1993 [31]
2. Provision of funding by your trading partner's top management for business process reengineering efforts in their firm required by your EDI connection with them
[5]
3. Strategic direction and business vision of trading partner Bhimji (interview); 1995; Hendrick and Ellram, 1993; Grover, 1990; La Londe and Cooper, 1989 [5,17,36,42] 4. Trading partner's openness to taking risks Hendrick and Ellram, 1993; Grover, 1990 [5,17,42] 5. Trading partner's fulfilment of the requirements of `just-in-time' La Londe and Cooper, 1989 [17,42]
6. Trading partner's readiness to adopt innovative information technologies just `to stay in the game'
[17,31,42]
7. Trading partner's willingness to trust and view your firm as a collaborator rather than an adversary
Hendrick and Ellram, 1993 [17,31,42]
8. Cross-functional orientation of your trading partner's internal working relationships, such as the extent to which cross-functional work teams and task forces are used
[5,17,31,36,42]
9. Assignment of valuable personnel by your trading partner's top management to the business process reengineering work for the EDI network
[5,17,36,42]
10. Your trading partner's fulfillment of the requirements of total quality management programs
Hendrick and Ellram, 1993 [17,31,42]
11. Use of participative rather than authoritative management and decision making style by your trading partners
Grover, 1990 [5,17,31,42]
12. Your trading partner wants and expects to have a long-term relationship with you
Hendrick and Ellram, 1993 [17]
13. Your trading partner's willingness to participate in joint committees to plan for and manage the EDI network
Hendrick and Ellram, 1993 [17,31,42]
14. Your trading partner's willingness to work with your firm in correcting or retransmitting data when previous EDI data transmissions are
inaccurate, untimely, or faulty
[17,42]
15. Your trading partner's fulfillment of the requirements of high customer service quality levels
Hendrick and Ellram, 1993; La Londe and Cooper, 1989 [17,42]
16. The extent to which your trading partner's top management envisions the business process reengineering effort for the EDI network as essential to their business strategy
Bhimji (interview), 1995; Barber (interview), 1995 [5,17,42]
17. Your trading partner's commitment to the goals of the EDI network Hendrick and Ellram, 1993 [5,17,42] 18. Your trading partner's rewards for business process reengineering efforts
related to the EDI network
[5]
19. Open and frequent communications between your trading partner and your firm
Hendrick and Ellram, 1993 [5,17,31,42]
20. Your trading partner's ability to send you timely, accurate, and complete information, which lets your firm serve your customers better
Hendrick and Ellram, 1993; La Londe and Cooper, 1989 [17,42]
21. EDI `readiness' of your trading partner: possession of technology (hardware/software) and expertise on EDI
Grover, 1990; La Londe and Cooper, 1989 [17,36,42]
22. Your trading partner properly assigns liability and responsibility in cases of errors in transmission, translation, storage, deletion, and destruction of data
[17]
23. Your trading partner's willingness to adapt to changes and reengineer business processes to achieve the goals of the EDI network
Barber (interview), 1995 [5,17,42]
24. Your trading partner's fulfillment of the requirements of `quick response'
[17,42]
25. Your trading partner's ability to respond to data transmissions on time La Londe and Cooper, 1989 [17,42] 26. Your trading partner's fulfillment of the requirements of
continuous improvement programs
re®ne the list, it was shown to several EDI managers, faculty members, and MIS Ph.D. students. They were asked to comment on the readability and completeness of the items. Their feedback led to the rewording of several items. These 31 items were made part of a questionnaire which also asked for additional information concerning the responding ®rm.
2.2. Sampling
Several ways were used to gather data. First, the support of the National Association of Purchasing Management (NAPM) was enlisted; they gave us access to their mailing list. A random sample of 2000 ®rms was selected from this list. The research packet (a copy of the questionnaire, self-addressed pre-paid envelope, and a cover letter explaining the purpose of the study) was mailed to the person in charge of EDI or electronic commerce in each ®rm. With the tendency towards downsizing in most infor-mation technology or management inforinfor-mation sys-tems shops regardless of company size, it was fairly reasonable to assume that the person in charge will be reasonably familiar with both the technical and busi-ness issues involved in an EDI network. A total of 140 questionnaires was returned resulting in a response rate of 7 percent. Additional questionnaires were distributed to prospective respondents at a national conference of the EDI professional association, and mailed to volunteers who responded to an advertise-ment inEDI World magazine. These sources netted 12 useable questionnaires resulting in a total of 152 responses. The seemingly low response rate raised
concerns about non-response bias and generalizability of results. However, a response rate between 5 and 10 percent for studies of this nature is considered normal [1].
When key characteristics of the population sampled are not known, a simple yet powerful method to check for non-response bias is the time trend extra-polation test [2,37]. A key assumption underlying this test is that non-respondents are more like late respondents than early respondents. Therefore, simi-larities in key variables between the early respondents and late respondents would indicate the absence of non-response bias. No statistically signi®cant dif-ferences were detected in ®rm size (number of employees), sales, industry type, and whether a ®rm is a customer or supplier in the two groups of respon-dents.
3. Results
3.1. Profile of firms
Table 2 shows the pro®les of the responding cus-tomer and supplier ®rms. The number of cuscus-tomer and supplier ®rms were almost even in the sample (77 and 75, respectively). More customer ®rms had over 10,000 employees (46.8 percent) than supplier ®rms (24 percent). In terms of net sales, a greater percentage of customer ®rms earned over a billion dollars as opposed to supplier ®rms. Eighty-one percent of the supplier ®rms were in manufacturing while about half (48 percent) of the customer ®rms were also in manufacturing.
Table 1 (Continued)
Criterion References
27. Your trading partner's readiness to adopt innovative information technologies to gain competitive advantage
Grover, 1990 [5,17,36,42]
28. Your trading partner's performance expectations and goals for the EDI network
Hendrick and Ellram, 1995 [5,17,31,42]
29. Your trading partner's willingness to handle exceptions and problems by negotiating
Hendrick and Ellram, 1993
30. Your trading partner's perception of the future of the industry and how their firm will position itself in response to this scenario
Grover, 1990; La Londe and Cooper, 1989 [17,42]
31. Existence of free-flowing and open communications within each of the participating firms, the buyer and the supplier
3.2. A factor analysis of trading partner selection criteria
Factor analysis is used to identify underlying con-structs from a large number of interrelated variables. Principal components analysis was used for factor extraction in order to obtain estimates of the initial factors that account for the largest variance in the sample. Table 3 shows the initial statistics generated for the six candidate factors. The rule used to ®nally determine the number of factors to include was Kai-ser's eigenvalue greater than 1 criterion [24]. This
resulted in a six-factor solution that explains 68 percent of the variation. Varimax rotation was chosen as the method for transforming the initial factors into a more meaningful con®guration.
A total of 31 items were factor analyzed. Stevens [45] recommended a cases-to-variables ratio of 5 : 1 to guarantee a reliable factor analysis procedure; how-ever, some researchers such as Fuller and Swanson [20] have worked with ratios as low as 2 : 1. There were a total of 152 cases for the 31 items, thus, resulting in a cases-to-variables ratio of roughly 5 : 1. This is more than satisfactory, considering the suggested ratio limits. Factor loadings resulting from the varimax rotation were evaluated using the thresh-old of 0.35 recommended by Churchill [13]. One item (#I7) had a factor loading of 0.34 (slightly below the recommended value of 0.35), but was kept as it was deemed to be an important item. After assembling the ®rst set of items under each factor based on the speci®ed factor loadings, the list was reevaluated to streamline the items and ensure that its content and meaning ®t the theme of each respective factor. An item that had the highest loading on a factor but did not quite match the theme of that factor was moved to another factor with which it had a better ®t in terms of content and where its loading was still fairly high. Repositioning items under each factor to ®t a thematic criterion justi®ed by theory is recommended by Sethi and King [41].
A careful examination of the items bonded with each of the six factors (Tables 4 and 5) leads us to name the six factors as follows:
Factor 1: Strategic commitment. Factor 2: Trading partner flexibility. Factor 3: Joint partnering for EDI. Factor 4: Readiness for high-level EDI. Factor 5: EDI infrastructure.
Factor 6: Communication.
In order to determine the degree of emphasis placed on the EDI partner selection criteria by customer and supplier ®rms, scores for the six factors were com-puted. Table 6 shows the mean and standard deviation of the six factors for the two groups of ®rms.
Note that both groups rated `Communication' and `EDI Infrastructure' as number one and two, respec-tively. To measure the extent to which the rankings given by the two groups agree, Kendall's coef®cient of
Table 2 Profile of firms
Customer firms (%) Supplier firms (%)
(n77) (n75)
No. of employees
Under 200 10.4 29.3
200±1000 9.10 18.7
1001±5000 22.1 20.0
5001±10,000 5.2 5.3
Over 10,0000 46.8 24.0
Not reported 6.5 2.7
Net sales (in millions)
Under 10 7.8 12.0
10±100 3.9 20.0
100±1000 18.2 18.7
Over 1 billion 61.0 38.7
Not reported 9.1 10.7
Firm type
Manufacturing 48.0 81.0
Service 12.0 2.0
Retailing 19.5 2.7
Other 11.5 5.2
Not reported 9.0 9.1
Table 3
Eigenvalues and percentage variation explained
Factor Eigenvaluea % Variance
1 12.29 38.4
concordance [35] was calculated. It showed that the rankings are not statistically signi®cantly different at the 0.05 level.
To determine whether the mean scores of factors for customer ®rms are different from those of supplier ®rms, a series oft-tests were conducted. First, note that the means for customer ®rms are higher than means for supplier ®rms in each of the six factors. The last two columns of Table 6 show the results of theset-tests. At the 0.05 level, statistically signi®cant differences were found for all factors except `joint partnering for EDI' and `strategic commitment'.
Table 7 reports the results when the level of EDI implementation (Level 1, 2, or 3) is taken into account. Among ®rms at Level 1, customer ®rms indicated a signi®cantly higher degree of emphasis on `commu-nication', `trading partner ¯exibility', and `readiness for high level EDI implementation' compared to supplier ®rms. At Level 2, the two groups differed on the latter two factors. At Level 3, differences existed on the last factor±`readiness for high level EDI implementation'.
Figs. 1 and 2 display the mean factor scores for customer and supplier ®rms, respectively, when bro-ken down by the level of EDI implementation. Using multivariate analysis of variance (MANOVA), it was found that the customer ®rms signi®cantly differed with respect to the six factors across the three levels of EDI implementation at the 0.05 level. Follow-up analysis using univariate F-tests revealed that the differences are signi®cant on Factor 3 (joint partner-ing for EDI) and Factor 5 (EDI infrastructure). Also, Level 1 means are statistically signi®cantly lower than the means for Level 2 and Level 3. Similar analysis for supplier ®rms showed that ®rms at Level 1 have signi®cantly lower means on the six factors than ®rms at levels 2 or 3.
3.3. Comparison of selection factors
In order to examine whether customer ®rms use different criteria than supplier ®rms, the scores for the six factors were compared for the two types of ®rms using two-samplet-tests. Table 6 shows these results for each of the six factors. Statistical signi®cant differences were found atp< 0.05 between customer and supplier ®rms on all factors except joint partnering for EDI and strategic commitment. The factor strate-gic commitment was, however, signi®cant atp< 0.10. It is worthwhile to note that the mean scores for customer ®rms are consistently higher than those for supplier ®rms for each factor, indicating that while selecting suppliers, customer ®rms emphasize these factors to a higher degree.
Differences between customer and supplier ®rms both at the overall and detailed EDI implementa-tion levels are greatest for the following factors: readiness for high-level EDI (p< 0.001), trading partner ¯exibility (p< 0.001), and communications (p < 0.01).
Table 4
Matrix of factor loadingsb
Itemsa 1 2 3 4 5 6
aThe following are the six congruence factors: 1: Top-level strategic commitment, 2: Trading partner flexibility, 3: Joint partnering for EDI, 4: Readiness for high-level EDI implementa-tion, 5: EDI infrastructure, and 6: Communications.
Table 5
Six trading partner selection factors
Variable
Corresponding questionnaire itema
Factor 1: Top-level strategic commitment
I3strategic direction and business vision of this trading partner; I6your trading partner's readiness to adopt innovative technologies just to `stay in the game'; I18the extent to which your trading partner's top management envisions the business process reengineering effort for the EDI network as essential to their business strategy; I19your trading partner's commitment to the goals of the EDI network; I20your trading partner's rewards for business process reengineering efforts related to the EDI network; I29your trading partner's readiness to adopt innovative information technologies to `gain a competitive advantage'; I30your trading partner's performance expectations and goals for the EDI network; I32your trading partner's perception of the future of the industry and how their firm will position itself in response to this scenario
Factor 2: Trading partner flexibility
I1your trading partner's flexibility and willingness to meet your information needs and other needs; I2provision of funding by your trading partner's top management for business process reengineering efforts in their firm required by your EDI connection with them; I7your trading partner's willingness to share information; I8your trading partner's willingness to take risks; I25your trading partner's willingness to adapt to changes and reengineer business processes to achieve the goals of the EDI network; I31your trading partner's willingness to handle exceptions and problems by negotiating
Factor 3: Joint partnering for EDI
I9your trading partner's willingness to trust and view your firm as a collaborator rather than an adversary; I10cross-functional orientation of your trading partner's internal working relationships, such as the extent to which work teams and task forces are used; I13use of participative (rather than authoritative) management and decision making style by your trading partner; I14your trading partner wants and expects to have a long-term relationship with you; I15your trading partner's willingness to participate in joint committees to plan for and manage the EDI network
Factor 4: Readiness for high-level EDI implementations
I5your trading partner's fulfillment of the requirements of `just-in-time'; I12your trading partner's fulfillment of the requirements of total quality management programs; I17your trading partner's fulfillment of the requirements of high customer service quality levels; I26your trading partner's fulfillment of the requirements of `quick response'; I28your trading partner's fulfillment of the requirements of continuous improvement programs
Factor 5: EDI Infrastructure
I11assignment of valuable personnel by your trading partner's top management to the business process reengineering work for the EDI network; I16your trading partner's willingness to work with your firm in correcting or retransmitting data when previous EDI data transmissions are inaccurate, untimely, or faulty; I23EDI `readiness' of your trading partner: possession of technology (hardware/software) and expertise on EDI
Factor 6: Communications
I21open and frequent communications between you and your trading partner; I22your trading partner's ability to send you timely, accurate, and complete information, which lets your firm serve your customers better; I27your trading partner's ability to respond to data transmissions on time; I33existence of free-flowing and open communications within each of the participating firms, the buyer and the supplier
aAll references to `trading partner' in the questionnaire item column was worded using either `customer' or `supplier', depending on the set of questionnaire being used. **I3, for instance, is to be read `item 3'.
Table 6
t-test results for customer and supplier firmsa
Factors Mean (Customer) Mean (Supplier) t p
(n74) (n78)
Communications 4.44 (0.56)b 4.12 (0.79) 2.83 0.005***
EDI infrastructure 4.15 (0.70) 3.88 (0.79) 2.29 0.023
Joint partnering for EDI 3.72 (0.77) 3.57 (0.88) 1.15 0.251
Strategic commitment 3.76 (0.64) 3.52 (0.92) 1.84 0.068*
Trading partner flexibility 3.92 (0.49) 3.38 (0.88) 4.76 0.000***
Readiness for `High Level' EDI implementation 4.02 (0.76) 3.28 (0.97) 5.26 0.000***
3.4. MANOVA results comparing trading partner selection factor means
One-way multiple analysis of variance (MANOVA) was also performed to determine if there are
signi®-cant differences between the two groups of ®rms, customers and suppliers, in their evaluation of the six trading partner selection factors, which are the dependent variables in this case. Customers consis-tently assigned higher mean values for the six factors
Table 7
Comparison of customer and supplier firms by implementation levela
Factor Level 1 Level 2 Level 3
Customer (n74)
Supplier (n78)
Customer (n74)
Supplier (n78)
Customer (n74)
Supplier (n78)
Communications 4.37* 3.90* 4.48 4.28 3.45 4.34
EDI infrastructure 3.98 3.71 4.27 3.99 4.33 4.05
Strategic commitment 3.62 3.35 3.84 3.51 3.92 3.98
Joint partnering for EDI 3.44 3.29 3.93 3.81 3.99 3.76
Trading partner flexibility 3.84* 3.10* 4.00* 3.47* 3.94 3.86
Readiness for high-level EDI 3.91* 3.02* 4.07* 3.42* 4.17* 3.63*
a*Indicates significant at the 0.05 level.
than did suppliers. MANOVA results reveal signi®cant differences between the two types of ®rms based on the Pillais, Hotellings, and Wilks tests, all with a signi®cance ofF-value of 0.000 (p< 0.001). It shows the results of the univariate F tests disclosing in greater detail the level of differences between the ®rms for each of the six factors.
Customers and suppliers agree on the importance of only one factor, joint partnering. Otherwise, there are signi®cant differences between the ®rms in their per-ception of the criticality of the remaining ®ve factors. Customers and suppliers are most distant in their evaluation of trading partner ¯exibility, readiness for high-level EDI, and communications (p< 0.01). The gap between the ®rms closes slightly in their evaluation of EDI infrastructure (p< 0.05) and is narrowed even more so in their assessment of strategic
commitment (p< 0.10). These MANOVA results
recon®rm thet-test results obtained.
4. Discussion of results
This study presents the organizational dimensions to consider when selecting EDI trading partners. Various trading partner selection criteria for EDI partnerships and a methodology for operationalizing its component factors, which focus on certain themes were presented. The customer ®rm, which is usually the hub ®rm, is more insistent about certain techno-logical initiatives that directly affect the EDI network. The importance of being selective is demonstrated by a study which found that the ®rms that were more discriminating in choosing their EDI trading partners were able to implement systems that have greater range, width, and depth [50].
Factor analysis yielded six factors in trading partner selection: strategic commitment, trading partner ¯ex-ibility, joint partnering for EDI, readiness for high-level EDI, EDI infrastructure, and communication. A
comparison of customer and supplier ®rms showed that both types of ®rms equally considered commu-nication and EDI infrastructure important in terms of
the ranking of means. A series of t-tests of the
differences of means were conducted to detect if customer ®rms expressed their preferences differently from supplier ®rms. Overall, customers had higher means for all factors than did suppliers. Both types of ®rms, though, were similar in their evaluation of all factors except joint partnering for EDI and strategic commitment at the 0.05 level.
An analysis of differences of the means was also conducted across the three levels of EDI implementa-tion. Using MANOVA, it was found that customer ®rms differed signi®cantly from supplier ®rms in their evaluation of the six factors across the three levels of EDI implementation at the 0.05 level. MANOVA
results con®rmed t-test results in ®nding out that
the differences between customer and supplier ®rms were greatest for readiness for high-level EDI, trading partner ¯exibility, and communications.
It makes sense that these two groups of ®rms should differ the most in their evaluation of these three factors which are more heavily involved in the high-level and long-term implementation of EDI. Customer ®rms in the sample, which are much larger both in terms of number of employees and net sales, assigned a higher level of criticality to the items of these factors. This could indicate actual involvement in higher level EDI implementation and/or a better conceptual apprecia-tion of the bene®ts from these initiatives. It usually takes experience, readiness to move up to more chal-lenging levels of implementation, or a sophisticated appreciation of the opportunities for more aggressive initiatives for strategic advantage to be able to respond to the factor items of these speci®c factors with a greater sense of urgency. Business executives of the more aggressive ®rms usually possess the savvy to employ new telecommunications-based technologies such as EDI to develop new strategies, to transform their organization, and to survive competitive threats. Elements of Factor 4, readiness for high-level EDI , suggest awareness of more advanced EDI initiatives such as `just-in-time', total quality management pro-grams, `quick response' in retailing [28], continuous improvement programs or vendor-managed inventory in manufacturing or services industries, and achieving high customer service quality levels.
5. Conclusions
5.1. Limitations of the study
The response rate, though acceptable, could still be improved in a future replication to strengthen the representativeness of the ®ndings. The sample size of 2000 ®rms to which the research packet was sent is only a small part of the relevant population. Financial and logistical constraints made it dif®cult to overcome this limitation. Also, the study assumed that the views of the head of the EDI or electronic commerce group of the ®rm were accurate and representative of those of the ®rm. Large EDI projects are usually undertaken in teams pulling together the talents of cross-functional members with multiple contributions and opinions. Data was not captured on the length of the relationship between the trading partners prior to their EDI con-nection. This experience may have had an effect on their views of what constitutes a good partner. This study looks only at the attributes of what makes a good EDI trading partner. That is a narrow area within a much larger domain investigating the concept of what makes a good trading partner under different organi-zational and technological conditions.
5.2. Managerial implications
While there is a tendency for powerful hubs to coerce potential trading partners into their EDI net-works, this is true, to a large extent, if the hub is dealing with much smaller ®rms. The aggressive stance may not be a very prudent approach to take, though, if the supplier under consideration is a much larger ®rm with valuable and scarce resources and therefore, has much more negotiating leverage. The following discussion points out more speci®c recom-mendations that may be more appropriate for medium-to large-sized cusmedium-tomer and supplier ®rms.
in collaborating with a potential hub ®rm and also, participate more effectively in undergoing the strategic, technological, and organizational changes required by EDI.
5.2.1. Suppliers
Suppliers need to recognize and prepare for `high-gear' performance that their customer counterparts expect. This means developing the attributes that make for trading partner ¯exibility, studying the requirements of certain level 3 EDI implementations, and facilitating open communications. Since suppli-ers, in effect, constitute a `direct appendage' to customer ®rms that may desire to move in the direc-tion of market leadership through high-level EDI, it behoves these suppliers to be able to shift its resources `on immediate notice'. `High-gear' perfor-mance refers to moving into level 3 EDI implementa-tions such as `just-in-time' manufacturing, vendor-managed inventories or continuous replenishment systems, `quick response' in retailing, `ef®cient con-sumer response' in the grocery and food distribution business, etc.
5.2.2. Customers
Customers may also need to underwrite the costs of being the `leader' trading partner in the dyad [23]. Customers that are in a position to be `hubs' and these
are usually the more prosperous and larger
®rms, should use the ®ndings of this study in more carefully selecting and assessing its potential trading partners. Obviously, suppliers that have the attributes that customers in this study considered critical should be considered ®rst in the short list of candidate partners. If the customer has to absolutely deal with certain suppliers that do not have the desired attri-butes, the customer could provide both educational and technical subsidy to help close the gap between itself and the supplier with the unique or much-needed resources.
In their study of small ®rms and the adoption of EDI, Iacovou, Benbasat, and Dexter [23] recommend that large EDI network hubs mount in¯uence strate-gies, based primarily on promotional efforts and noncoercive tactics. Since EDI partnerships are ide-ally cooperative endeavors, resorting to threats and punishment basically amounts to sabotage and would be counterproductive in the long run. An effective
customer hub ®rm could learn from the ®ndings of this study by designing educational campaigns that teach potential trading partners not only the bene®ts of EDI but also the components that will best prepare future trading partners especially for appreciating the elements of trading partner ¯exibility, readiness for high-level EDI, and communications. While powerful large hubs still tend to coerce smaller ®rms into their EDI network, this approach is inappropriate and risky when applied to much larger potential trading partners with scarce resources that are much needed for long-term, collaborative partnering arrangements.
References
[1] P.L. Alreck, R.B. Settle, in: The Survey Research Handbook, Homewood, IL, Richard D. Irwin (Publisher), 1985. [2] J.S. Armstrong, T.S. Overton, Estimating non-response bias in
mail surveys, Journal of Marketing Research 14 (3), 1977, pp. 396±402.
[3] J.Y. Bakos, E. Brynjolfsson, Information technology, incen-tives and the optimal number of suppliers, Journal of Management Information Systems 10 (2), 1993, pp. 37± 53.
[4] J.Y. Bakos, A strategic analysis of electronic marketplaces, MIS Quarterly 15 (3), 1991, pp. 295±310.
[5] N. Barber, The Organizational Aspects of EDI: A Project Manager's Guide, Electronic Data Interchange Association Alexandria, VA, 1989.
[6] S. Barrett, B.R. Konsynski, Inter-organizational information sharing systems, MIS Quarterly 6, 1982, pp. 93±105. [7] F. Bergeron, L. Raymond, Managing EDI for corporate
advantage: a longitudinal study, Information & Management 31 (6), 319±333.
[8] R.D. Buzzell, G. Ortmeyer, Channel partnerships streamline distribution, Sloan Management Review 36 (3), 1995, pp. 85± 96.
[9] J.I. Cash, J.I., B. Konsynski, IS Redraws Competitive boundaries, Harvard Business Review, 1995, 134±142. [10] J. Cathey, What a controller should know, Management
Accounting, 1992, pp. 47±51.
[11] W.G. Chismar, J. Meier, A model of competing interorgani-zational systems and its application to airline reservation systems, Decision Support Systems 8 (5), 1992, pp. 447± 458.
[12] V. Choudhury, Sustaining Competitive Advantage with Interorganizational Information Systems, in: Proceedings of HICSS-21, IEEE Computer Society Press, Washington, DC 1988, pp. 44±51.
[14] M.C. Cooper, L.M. Ellram, J. Gardner, A.M. Hanks, Meshing multiple alliances, Journal of Business Logistics 18 (1), 1997, pp. 67±89.
[15] C.W. Crook, R.L. Kumar, Electronic data interchange: a multi-industry investigation using grounded theory, Informa-tion & Management 34, 1998, pp. 75±89.
[16] M.A. Emmelhainz, EDI: A Total Management Guide, Van Nostrand Reinhold, New York, 1990.
[17] M.A. Emmelhainz, EDI: A Total Management Guide, Van Nostrand Reinhold, New York, 1993.
[18] D.M. Ferguson, N.C. Hill, J.V. Hansen, Electronic data interchange: foundations and survey evidence on current use, Journal of Information Systems, 1990, pp. 81±91.
[19] S.S. Fiorito, E.G. May, K. Straughn, Quick response in retailing: components and implementation, International Journal of Retail & Distribution Management 23 (5), 1995, pp. 12±21.
[20] A.L. Fuller, E.G. Swanson, Information centers as organiza-tional innovation, Journal of Management Information Systems 9 (1), 1992, pp. 47±68.
[21] J.V. Hansen, N.C. Hill, Control and audit of electronic data interchange, MIS Quarterly 13 (4), 1989, pp. 403±413. [22] M. Holmlund, S. Kock, Buyer-dominated relationships in a
supply chain: A case study of four small-sized suppliers, International Small Business Journal 15 (1), 1996, pp. 26± 40.
[23] C.L. Iacovou, I. Benbasat, A.S. Dexter, Electronic data interchange and small organizations: adoption and impact of technology, MIS Quarterly 19 (4), 1995, pp. 465±481. [24] H.F. Kaiser, An index of factorial simplicity, Psychometrika
39, 1974, pp. 31±36.
[25] L.A. Kappelman, T.C. Richards, R.J. Tsai, Conducting business on the information highway: a manager's guide to electronic data interchange, Business Forum 20 (3±4), 1995, pp. 29±32.
[26] K. Kumar, H.G. van Dissel, Sustainable collaboration: managing conflict and cooperation in interorganizational systems, MIS Quarterly 20 (3), 1996, pp. 279±300. [27] R. Landeros, R. Monczka, Cooperative buyer/seller
relation-ships and a firm's competitive posture, Journal of Purchasing & Materials Management 25 (3), 1989, pp. 9±18.
[28] W. Lane, T. Sample, U.M. Apte, W.E. Vaughan, Quick Response and EDI at JCPenney and Haggar Apparel: Using Information Technology for Re-engineering of Business Processes, Working Paper, Cox School of Business, Southern Methodist University, Dallas, TX, May 1993.
[29] M. Manheim, in: L.M. Harasim (Ed.), Integrating Global Organizations through Task/Team Support Systems, in Global Networks: Computers and International Communication, Cambridge, MA, 1993.
[30] J. Meier, The importance of relationship management in establishing successful interorganizational systems, Journal of Strategic Information Systems 4 (2), 1995, pp. 135±148.
[31] P. Moody, Breakthrough Partnering: Creating a Collective Enterprise Advantage, Oliver Wright Publications, Essex Junction, VT, 1993.
[32] E.A. Murray, J.F. Mahon, Strategic alliances: gateway to the new Europe? Long Range Planning 26 (4), 1993, pp. 102± 111.
[33] S. Neumann, Strategic Information Systems: Competition Through Information Technologies, Macmillan College Pub-lishing, New York, 1994.
[34] S.R. Nidumolu, Interorganizational information systems and the structure and climate of seller-buyer relationships, Information & Management 28, 1995, pp. 89±105. [35] M.J. Norusis, SPSS Base System User's Guide, SPSS,
Chicago, IL, 1990.
[36] D. Petrosky, M. Emmelhainz, Selecting EDI Trading Partners, EDIWORLD 1 (6) (1991) 24±25, p. 34.
[37] L.W. Phillips, Assessing measurement in key information reports: a methodological note on organizational analysis in marketing, Journal of Marketing Research 18, 1981, pp. 395± 415.
[38] F.J. Riggins, C.H. Kriebel, T. Mukhopadhyay, The growth of interorganizational systems in the presence of network externalities, Management Science 40 (8), 1994, pp. 984± 998.
[39] F.J. Riggins, T. Mukhopadhyay, Interdependent benefits from interorganizational systems: opportunities for business part-ner reengineering, Journal of Management Information Systems 11 (2), 1994, pp. 37±57.
[40] J.A. Senn, Electronic data interchange: the elements of implementation, Information Systems Management 9 (1), 1992, pp. 45±53.
[41] V. Sethi, W.R. King, Construct measurement in information systems research: an illustration in strategic systems, Decision Sciences 22, 1991, pp. 455±472.
[42] P.K. Sokol, From EDI to Electronic Commerce: A Business Initiative, McGraw-Hill, New York, 1995.
[43] R.H. Sprague, Jr., B.C. McNurlin, Information Systems Management in Practice, Prentice-Hall, Englewood Cliffs, NJ, 1993.
[44] W. Stallings, Business Data Communications, Macmillan, New York, 1990.
[45] J. Stevens, Applied multivariate statistics for the social sciences, Lawrence Erlbaum, Hillsdale, NJ, 1986.
[46] R. Suomi, Inter-organizational information systems as company resources, Information & Management 15, 1988, pp. 105±112.
[47] L.R. Viyayasarathy, D. Robey, The effect of EDI on market channel relationships in retailing, Information & Management 33, 1997, pp. 73±86.
[48] R.P. Vlosky, P.M. Smith, D.T. Wilson, Electronic data interchange implementation strategies: A case study, Journal of Business & Industrial Marketing 9 (4), 1994, pp. 5±18.
[49] E.T.G. Wang, A. Seidmann, Electronic data interchange: competitive externalities and strategic implementation poli-cies, Management Science 41 (3), 1995, pp. 401±418. [50] L.R. Williams, G.D. Magee, Y. Suzuki, A multidimensional
Rebecca Angelesis Assistant Professor at the Information and Decision Sciences Department, School of Business, Montclair State University, New Jersey. Her research publications have appeared in such publications as
Decision Support Systems,International Journal of Physical Distribution & Logistics Management, Pricing Strategy and Practice: An International Journal,and EDI Forum: The Journal of Electronic Commerce.
Her research interests are in the areas of electronic commerce, electronic trading partnerships, electronic data interchange (EDI), Internet-EDI, and Interorganizational systems, and
innovative education approaches in Management Information Systems.
