STX4TEGY AN EMPIRICAL ANALYSIS OF BENEFITS OF ELECTRONIC DATA INTERCHANGE (EDI) IMPLEMENTATION: IMPLICATIONS FOR NEW IT IMPLEMENTATION Deepak Khazanchi University of Nebraska at Omaha khazanchi unomaha. edu ABSTRACT This paper reports that the benefits accrued from implementing and integrating Electronic Data Interchange (EDI) within small and medium-sized enterprises (SMEs) can be conceptualized into two factors. First, Jirms derive operationalltactical benefits by predominantly focusing on increasing internal utility of this technology. Second, firms derive strategic benefits Pom EDI in the form of beuer external relationships and alliances with trading partners and an enhanced ability to compete in their market. Among oiher significant findings, there are clear indications Pom the correlation statistics reported here that experience with EDI, industrial category of a Jirm and the level of EDI integration have a significant influence on the ability of a firm to obtain long-term (strategic) bene/its from such IT projects. These results also have significant implications for SME managerslstakeholders considering new interorganizational ITinitiatives. INTRODUCTION According to forecasts published by Giga, a private research firm, Electronic Data Interchange (EDI) transactions in the United States alone were about $2.7 billion in 1997 and are estimated to grow to $3.8 billion by 2002 (Wilson, 2000). Vollmer (2001), a research director of 82B integration at Giga Information Group asserts: "During the past several years, it has been all too common to hear "experts" denigrate the potential of EDI in favor of some new solutions just around the corner. However, it is no coincidence that both ebXML and BizTalk Server- leading XML-based initiatives to build widespread e-business functionality-are supporting existing EDI transactions. Alter a lengthy review of available options, the sponsoring organizations came to the same conclusion- EDI is the only practical e-business standard that makes sense for basic functionality at this time." The Giga report challenges the popular notion that traditional EDI transactions will be widely replaced by emerging Web alternatives. In supporting the analysis presented in this report, 45 Journal ofSmall Business Strategy Vo/. l3, No. l Spring/Summer 2002 Jack Reich, the E-commerce director of National Gypsum states: "I don't know of a single company in our industry, for all the hoopla, that's exchanging documents via the Web in XML format... Many of the large players have used traditional electronic data interchange for years. We'e had good success with EDI over the Web as a cost-saving alternative.'" Clearly, though a shrinking percentage of the total business-to-business (B2B) electronic commerce pie, EDI continues to be "alive and kicking" and an important element of the future landscape of global B2B e-commerce (Ibid.). Electronic Data Interchange Flectronic Data Interchange (EDI) is the computer-to-computer interchange of business transactions that conforms to specified standards over a communications network that includes at least two trading partners. These interactions include the interchange of common commercial information typically consisting of purchase orders, shipping notices, invoices, related acknowledgements, funds transfer with banks, etc. (Zorfass Et Michel, 1992). EDI automates the slow, labor-intensive exchanging of transactional documents in paper form via fax and/or regular mail. The EDI enterprise is the hub of activities. Hubs represent the accumulation point for transactions from multiple trading partners. For example, Wal-Mart is a hub with more than 5000 electronic hook ups with its vendors. The trading partners can be viewed as spokes. Spokes (vendors, customers, etc.) become part of the extended EDI enterprise. Larger spokes can be hubs of their own supplier, customer networks. Most SMEs tend to be spokes for large hub organizations. EDI requires five key elements (Arunachalam, 1995; Pfeiffer, 1992): Electronic mail for rapid personal (administrative) communications; On-line networks for rapid communications such as third party or value added networks (VANs) and Virtual private networks (VPNs); At least two organizations conducting joint business transactions electronically (trading partners); Standard protocols for file and message transfers. This is accomplished with trading partner agreements regarding data coding and formatting rules. Standard EDI message formats can be those developed by industrial organizations (e.g., TDCC/EDIA, VICS, WINS), proprietary (e.g., General Motors), national (ANSI X12) or International (UN/EDIFACT). Data processing task(s) at both (all) organizations pertaining to a transaction are supported by independent application systems. There are three generic approaches to implementing EDI links. The first approach uses a direct EDI link between vendor and customer using a modem and telephone line. Many large hub organizations own and operate a private network service (e.g., Wal-Mart, GE) that all business partners are required to use. Trading partners establish communications using a dial- up link to the hub's network. While a majority of these hubs do not charge for their network service, trading partners do have to pay all phone charges. The second approach revolves around indirect EDI links through value-added networks (VAN) or "third party electronic clearing houses." These independent EDI networking vendors provide all the necessary soltware and communications services and essentially perform the function of an electronic post office for numerous business partners. Trading partners place their business documents in "electronic envelopes" identifying the sender and receiver. The document is mailed to the VAN aller setting up a dial-up link via phone lines. The VAN will either forward the document to the hub organization's computer automatically or place it in the receiver's mailbox for pickup at a later time. Major costs associated with this EDI transmission option 'EDI in XML Envelope", http;//www.intemetwk.corn/(Apr 23', 2001). 46 Journal of Small Business Strategy Vo/. /3, No. I Spring/Summer 2002 will include expenses relating to VAN setup, telephone lines, and monthly transaction fees. Third, with the development of better Internet browsers and compatible EDI software that incorporates adequate security measures including encryption, the robust and cheaper Internet has become the medium of choice for transmitting electronic documents and messages globally. This approach is essentially similar to the direct communications link except that the Internet access charges are substantially lower than the other options. RESEARCH RATIONALE AND QUESTIONS New business practices such as lust-In-Time (JIT) manufacturing and quick response retailing (QR) rely on the transfer of transaction data to gain a competitive advantage in the market place. Pickett and Udo (1994) assert that "the numerous benefits of doing business using electronic data interchange (EDI) have caused large companies to accept EDI as a way of life." In a longitudinal study of Chrysler's adoption of EDI, Mukhopadhyay, Kekre, and Kalathur (1995) report substantial dollar savings due to improved information exchanges between Chrysler and its suppliers that result from EDI. They also assert that, not unlike many major hub enterprises, Chrysler made EDI a necessary condition for suppliers doing business with their assembly centers. However, most small companies at the receiving end of this EDI mandate do not take complete advantage of this strategic technology through appropriate consideration of costs/benefits and internal and external integration (Khazanchi, 1995). Furthermore, some research studies have found that businesses (small or large) that voluntarily initiate EDI have better success integrating it within internal functions and consequently realizing both operational and strategic benefits (Raymond & Bergeron, 1996; Swatman, Swatman & Fowler, 1994; Swatman & Swatman, 1991). For EDI to be a successful and eITicient means of electronic trading, whatever ultimate form the technology itself takes, a better understanding of the business impact of EDI and similar interorganizational information systems on small- to medium-sized enterprises (SMEs) is essential. Impact of EDI on Organizations Impact of ED/ refers to the actual benefits EDI adopters receive from utilizing EDI. EDI benefits can be categorized into Indirect and Direct Benefits (lacovou, Benbasat, & Dexter, 1995). The following definitions are culled from the work of Pfeiffer (1992), Banerjee and Golhar (1993), Swatman et al. (1994), lacovou et al. (1995), Arunachalam (1995), and Benjamin, de Long, and Michael, (1990). indirect benefits such as improved customer service, increased operational efiiciency, improved trading partner relationships, and increased competitiveness are obtained by organizations that are proactive, have excellent organizational support, and their business applications are seamlessly integrated with EDI. Essentially such organizations view EDI as a strategic technology and a necessary tool for doing business. Whereas, direct benefits such as higher quality of information, reduced transaction costs, improved cash flows, and reduced inventory levels are obtained by organizations in the form of financial savings as a result of EDI adoption. This description of indirect and direct benefits is consistent with the conclusion reached by authors such as Cash and Konsynski (1985), Porter (1985), Porter and Millar (1985), Malone, Vates, and Benjamin (1987), Johnston and Vitale (1988), and Benjamin, et al. (1990) that information in general and interorganizational information systems such as EDI in particular have allowed some firms to improve operational eITiciency and coordination with trading partners and create and sustain a significant competitive advantage in the marketplace. 47 Journal ofSmall Business Strategy Val. /3, iVa. l Spring/Summer 2002 Table I: Summary of Key EDI - SME Research Studies AuthorlWork Raymond and EDI World lacovou et al. Carter et al. Institute (/995) (l995) ()982) Assess success Assess education Factors Research Impact of EDI on influencing EDI factors that have & training needs obtain advantages f FDI implementation from EDI Organizational context (organizational sUpport, Perceived implementation benefits, Research process, & Organizational y 'l ( ) co trol N/A readiness, N/Aprocedures), External pressure, integration level, EDI adoption & imposition level, integration EDI ED I advantages impact (operational, managerial, strategic) Descriptive study Research Field Multiple-case using survey & Design study/Survey Survey study follow-up interviews 500 Canadian SMEs from Manufacturing Sampling various sectors in Quebec and A . Small firms in 25 firms fromFrame American, Ontario provinces F A I Canada various sectors (149 responses) vrtth & 250 employees) Conclusions: Use SMEs had &=100 Pre-adoption education & training to Gain employees with awareness of EDI Quahty of $ 1-$20 million in benefits is low; commitment of organizational EDI system; annual sales; 9 of Overall readiness context of EDI is 10 I d f 0 f training programs10 implemented of small firms is crucial to the EDI at the request low; Need for should be attainment of designed to suit (e b f L of a cu st om e r; fi n an ci aIbenefits; Low Over half technological, intended Research imposition level audience; increased managerial Findings (voluntary understanding revenues & support; Small adoption) has a f .. f 'xpected (actual)profits; Majority firms are significant impact reduced reluctant to operational on the quality impediments; document cycle integrate EDI into organizational education & time, improved operations context. training must be accuracy of because of high state of the art &. information costs. begin early in the adoption process. 48 1 Journal ofSmall Business Strategy Vol. /3, /Vo. l Spring/Summer 2002 Impact of EDI on SMEs Small firms that have a favorable organizational context (i.e., top management support, personnel training, collaboration between functional areas, etc.) and are seeking to achieve high-levels of internal and external integration have a better chance of obtaining many operational and strategic benefits of EDI. A summary of key research on the experience of SMEs with EDI implementation is provided in Table l. The results of EDI impact studies on SMEs are clearly indicative of the fact that small businesses can potentially accrue the same level of benefits as large firms if the following conditions are satisfied. In addition to being proactive to the changes in the business- technology environment, a business must have adequate organizational support, some degree of technological sophistication, adequate planning mechanisms in place, a sustained plan for the internal and external integration of EDI, and an awareness of the potential impact of EDI (i.e., direct and indirect benefits). Research Questions Based on the previous discussion, the purpose of this paper is to investigate two main questions. First, what is the nature and structure of benefits obtained by SMEs through EDI implementation? Second, what is the influence of various firm demographics and other variables (elicited from a priori research) on the EDI benefits construct? To address the latter question, the influence of variables such as "extent of trading partner support", "stage of EDI integration," "volume of EDI documents (messages)", "nature of cost/benefit analysis", "perceived benefits of EDI", etc. on EDI benefits is evaluated. These variables were previously identified by various researchers (e.g., Carter, Monczka, Clauson, & Zelinski, 1987; Monczka &. Carter, 1988; Pfeiffer, 1992; Swatman & Swatman, 1991; lacovou, et al., 1995) as having an impact on EDI adoption and integration and in consequence on the ability to realize potential benefits of EDI implementation. RESEARCH METHOD In the context of a larger study of the impact of EDI on SMEs in the Commonwealth of Kentucky, the previous two research questions were also addressed. The sampling frame was limited to Kentucky because this project was partially supported by a grant from the Kentucky Cabinet for Economic Development. A survey research design was used to elicit data about SMEs and their experiences with EDI implementation. Data Collection In late 1997, the survey was mailed to 353 SME-capable firms identified from the 1997 ED/ Yellow Pages (Phillips Business Information, lnc., 1997) and from two local hub companies and a State Government agency. Since high non-response rate'&60%) can dilute the ability to statistically generalize to the larger EDI user population, various measures to reduce nonresponse rates were taken resulting in an effective response rate of 24.3% or 86 useful responses. However, McDaniel and Gates (1993)report that higher response rates are a means to reducing nonresponse bias. They also report that "...ofall the studies that have looked for differences between nonrespondents and respondents (or early or later respondents) of mail surveys, none has been reported that found meaningful, practical differences between 't is well established that the possibility of a high non-response rate is a major problem with questionnaires (Sproull, 1988). 49 Journal ofSmall Business Siratcgv Vol. 13, Na. l Spring/Summer 2002 respondents and the entire sample or between early respondents and respondents as a whole" (pp. 233, emphasis added). Instrumentation EDl impact was measured in terms of relative benefits realized by SMEs through EDI adoption and integration. for convenience and readability of the survey instrument, EDI benefits were initially categorized into indirect and direct benefits as explained in the previous section of this paper. EDI Benefits Realized by Responding Firms In order to explore the nature of benefits realized by Kentucky SMEs, responding firms were asked to assess the impact of EDI implementation on their organization by indicating the extent io which each listed benefit had been obtained by the firm (refer column I, Table 2 for the list of items). Thus, responding firms rated the extent to which various benefits were obtained by their enterprise'. This was assessed with a 5-point Likert-type scale with verbal labels ranging from a score of I, "substantially deteriorated (or decreased)," to 3 or "no change," to 5, "substantially improved (or increased)." Thus checking a 5 would indicate that a firm had obtained a substantial improvement (or increase) in a specified benefit because of EDI implementation, whereas checking a I would indicate that a finn had observed a substantial deterioration (or decrease) in a specified benefit item. RESULTS AND DISCUSSION Profile of Survey Participants Industrial Sector and Range of Products All 86 responding firms provided information about their industrial sector. Table 2 profiles the sampled-firms by industry category. Table 2: Industry Category (N=86) Industry Category Frequency % of Responses Manufacturing 49 57% Wholesale Trade 23 27% Retail Trade 6 7% Services (e.g., computer, accounting, TV repair) 2 2% Transportation and Public Utilities I 1% Mining I 1% Other 4 In the manufacturing sector, participating firms make a diverse range of products including everything from industrial parts and supplies to candy and cheesecakes. In the wholesale trade sector, firms deal in products ranging from industrial parts and supplies to food and pharmaceuticals. The remaining firms are involved in retail trade such as office furniture and The exact phrasing of the question was as follows: Please evaluate the impact of EDI implementation in your organization by indicating the extent to which each of the following benefits have been obtained by your enterprise. Select a response by assessing the ~chan e observed in the listed EDI benefit. 50 Journal ofSmall Business Strategy Va/. /3, Na. l Spring/Summer 2002 power tools, services such as health and lab analysis, and other business activities such as hauling freight, warehousing, logistics management, and computer systems value added reselling (VAR). Respondent's Position (Job Title) An equal number (43) of responding individuals belong to the non-technical, managerial or administrative ranks (e.g., Treasurer, Owner/Major Stakeholder/President, Business Manager/General Manager) as those from the information systems branch (e.g., EDI Specialist/EDI Supervisor, IS Manager/EC Manager, Systems Analyst) completed the survey for sample-organizations. Organizational Size The sample is uniformly distributed by organizational size when size is measured in terms of the "number of full-time employees" (as displayed in Table 3). Table 3: Number of Full-Time Employees (N=86) Number of full-time employees Frequency % of Responses Fewer than 5 employees 5 6% 5 to 10 10 12% 11 to 20 7 8% 21 to 50 9 10% 51 to 100 11 13% 101 to 250 24 28% 251 to 500 7 8% More than 500 13 15% Another popular measure of organizational size is "sales volume" and is displayed in Table 4 below. A large number (nearly 70%) of responding firms had gross sales over $ 1 million in 1997 with more than half (47%) generating over $ 10 million in sales. Table 4: Estimated 1997 Gross Sales (N=86) 1997 Gross sales (Estimated) Frequency % of Responses Less than $ 10,000 I 1% $ 10,000 to $50,000 3 3% $50,001 to $ 100,000 4 $250,001 to $500,000 3 3% $500,001 to $ 1 million 3 3% $ 1 million to $5 million 12 14% $5 million to $ 10 million 8 9% More than $ 10 million 40 47% Don't know 12 14% Total 86 100% EDI Experience The utility a firm draws from EDI can also be gauged by its relationship with the length of EDI utilization or amount of experience gathered with this technology. Organizations with 51 Journal ofSmall Business Srrareqt Vol. /3, No. l Spring/Summer 2002 EDI experience of less than or equal to 12 months, make up nearly 10% of the sample, while 56% of the sampled-firms have more than one year and less than 5 years experience. Finally, organizations with more than 5 years of experience make up nearly 34% of the sample. Apparently, a majority of the firms in the sample report being substantially experienced with ED I. EDI Benellts Descriptive Analysis of EDI Benefits Realized by Responding Firms Each one of the benefits listed in Table 5 is significantly different from the middle scale value of 3.00 ("no change") when a one-sample t-test was applied at the 95% confidence level. In other words, on the average, survey-respondents reported achieving a small but statistically significant positive change in each of the listed benefits due to the implementation of EDI in their organization. It should be noted that "inventory levels" and "transaction costs" are reverse-coded and therefore, a deterioration (or decrease) in them has a positive influence on realized benefits. As shown in Table 5, the mean scores for all the individual EDI benefits clearly support this conclusion. However, it is surprising to note that none of the listed benefit categories has a mean score that falls in the slightly to substantially improved (or increased) or slightly to substantially deteriorated (or decreased) range'. Of course, there are individual firms in the sample that report having achieved substantial benefits from EDI, but on the average this is obviously not true. Table 5: Change in EDI Benefits-Descriptive Statistics (N=78) Potential EDI Benefits Realized'ean Standard Deviation Quality of Information 3.83 .93 Relationship with Trading Partners 3.83 .80 Customer Service 3.61 .80 Ability to Compete 3.59 .70 Operational Efficiency 3.55 .89 Cash Flows 3.33 .70 Transaction Costs (reverse coded ) 2.69 1.04 Inventory Levels (reverse coded ) 2.76 .51 Factor Analysis of EDI Benefits The "EDI benefits" items were further analyzed using the data reduction technique of "principal components analysis (varimax rotation with Kaiser Normalization)." This 'hese are equivalent to the ratings of 4 and 5 on the five point Likert-type "benefits" scale. 'espondents were asked to assess the impact of EDI implementation on their organization by indicating the extent to which each listed benefit had been obtained by the firm. A 5 point Likert-type scale with verbal labels was used with respondents indicating whether a benefit had "substantially deteriorated or decreased" (coded as a I), "slightly deteriorated or decreased" (coded as a 2), "no change" (coded as a 3), "slightly improved or increased" (coded as a 4), and "substantially improved or increased (coded as a 5). 6 Lowering transaction costs or reducing inventory levels has a positive impact on accruing benefits from EDI implementation. 52 Journal of Small Business Strategy Va/. /3, Na. l Spring/Summer 2002 exploratory factor analysis was used to identify any underlying factors that constitute the "EDI benefits" construct and for further understanding its relationship with previously identified variables. A two-factor structure was found, explaining nearly 58% of the sample variance. All the "EDI benefits" scale items had a loading greater than 0.5 on the factor to which they were attributed. Nunnally (1978) recommends a 0.5 threshold to achieve an adequate level of reliability for each factor in exploratory work. Communalities for the two factors range from 0.51 to 0.69 with one exception at 0.39. This result is another strong indication of the validity of the latent factor structure. Table 6: Factor Loadings for 'EDI Benefits'onstruct'perational/ Strategic Potential EDI Benefits Realized Tactical Benefits Benefits (BENEE/TJ) (BENEF/T2) Cash Flows (e.g., Improve cash flows by faster processing and .74 exchange of information between trading partners) Inventory Levels (e.g., Reduce inventory levels by shortening -.71 order cycle, reducing ordering costs) Operational Efficiency (e.g., Reduce lead time and costs, .64 .44 better information management, avoid re-keying of data) Transaction Costs (e.g., Lower costs by eliminating -.62 paperwork, postage, faxing, and saving on labor) Customer Service (e.g., Improve customer service by shorter .62 .51 lead times, timely information regarding transaction status) Quality of Information (e.g., Improve quality by increasing .57 .44 timeliness, accuracy, and accessibility of information) Ability to Compete (e.g., Increase ability to reach new .82 markets, provide better service at lower costs) Relationship with Trading Partners (e.g., Enhance trust by .81 sharing information, reduce errors, enable JIT/QR programs) Elgenvalues 2.59 2.04 % of Total Variance Explained (Cumulative) 32.41% 57 91% The two categories of potential EDI benefits realized by the surveyed organizations found by the factor analysis shown in Table 6 can be conceptually described as follows. Factor I can be named "operational/tactical benefits," and it relates to the change in benefits associated with the impact of EDI in engendering improved cash flows, reduced inventory levels, increased operational efficiency, lowering transaction costs, and improving quality of information. Factor 2 can be named "strategic benefits," and it relates to the change in benefits associated with the impact of EDI in increasing a firms'bility to compete and enhancing relationships with trading partners. As noted previously in the background section of this paper lacovou et al. (1995) categorized EDI benefits obtained by SMEs into indirect and direct benefits. They supported their conceptualization with seven case studies. The factor analysis reported above is based on a 'PSS/PC version 8.0 was utilized for statistical analysis. 'otation converged in 3 iterations. The extraction method used was Principal Component Analysis and varimax rotation with Kaiser normalization. Cross-loadings between factors below 0.25 are not shown. 53 Jottrna/ of Small Business Stra/eI0 Vo/. /3, No. / Spring/Summer 2002 sample size of 86" and clearly does not support their conceptualization of EDI benefit categories. Notwithstanding this finding, the individual benefit items identified from a priori research by lacovou et al. are useful indicants of the EDI benefits construct. Table 7: Relationship of SME Characteristics & EDI Benelitsnea (N=78) Industry EDI 1997 Gross ¹ Full-Time ¹ Temporary or Part-Time Category Experience Sales (Est.) Employees Employees Operational/ Tactical .120 .123 .036 .122 .098 Benefits (.294) (.285) (.756) (.289) (.393) (BENEFIT/) '""'gic 271v .348sv 055 .150 .112 (.017) (.002) (.633) (.191) (.328) Relationship of SME characteristics with EDI Benefits Table 7 summarizes the correlation statistics between variables that are useful in classifying surveyed-SMEs and the two EDI Benefit factors derived in the previous section of the paper. The data is consistent with the notion that firms with EDI experience can obtain greater strategic benefits from EDI implementation. On the other hand, experience with EDI does not seem to have a significant relationship with operational/tactical EDI benefits or benefits that accrue from improving the efficiency of internal operations and reducing cash flows. Interestingly, data analysis shows that the 'EDI benefits'ealized by SMEs are not significantly related to firm size measured in terms of either gross sales or number of employees (full or part-time). Finally, although industrial sector of sample firms has no significant relationship with the ability of a firm to obtain operational/tactical benefits from EDI implementation, it is significantly related with a firm's potential to realize strategic benefits from EDI implementation. Relationship of Other Key Variables to 'EDI Benefits'actors A number of other research variables have previously been identified by various researchers (e.g., Pfeiffer, 1992; Swatman & Swatman, 1991; Premkumar, Ramamurthy, & Nilakanta, 1994; lacovou, et al., 1995) as having an impact on EDI adoption and integration and in consequence on the ability of firms to realize potential benefits of EDI implementation. Table 8, 9 and 10 summarize the correlation statistics between these research variables and the two "EDI Benefits" factors. Actually the factor model was derived using 78 cases only. The remainder where excluded because of missing values." "Correlation is significant at the 0.01 level (2-tailed; 99% confidence); a Correlation is significant at the 0.05 level (2-tailed; 95% confidence)." Pearson correlation coeflicients with significance levels in parenthesis are shown. "These variables have previously been identified by various researchers (e.g., Pfeiffer, 1992; Swatman & Swatman, 1993) and lacovou et al., 1995) as having an impact on EDI adoption and integration and in consequence on the ability to realize potential benefits of EDI implementation. 54 Journal ofSmall Business Strategy Vol. /3, No. I Spring/Summer 2002 Table 8: Relationship of Key Determinants of EDI Benefits with OperationaUTactical Benefits Factor (Benefitl) and Strategic Benefits Factor (Beneflt2)" /Vature of Volume of EDI CosVBenefit Analysis Documents Current Stage of by SME Prior to (Messages) EDI Integration Adopting EDI Exchanged'earson BENEFIT I Correlation .261* .122 .456sa Sig. (2-tail) .022 .157 .000 BENEFIT I Correlation .108 200x .210*'a Sig. (2-tail) .348 .020 .074 N 77 78 78 Extent of Trading Partner support Hard- Soft- duration d Telecommuni- Impiemrn-Maintenance ware ware Training cation casts rattail BENEFIT ..188ass .336as .209sss .148 .127 .138 Correlation 'ig.(2-tail) .103 .003 .066 .197 .268 .138 BENEFIT M 0 026 102 055 002 017 079 Correlation Sig. (2-tail) .824 .374 .634 .984 .885 .491 N 76 78 78 78 78 78 Nature of Cost/Benefit Analysis Conducting a cost-benefit analysis prior to EDI implementation could provide firms an appreciation of whether EDI would be advantageous to them while understanding its inherent costs. Survey respondents were asked to identify the nature of cost/benefit analysis conducted by them prior to adopting EDI. A majority of the responding firms did not conduct any cost/benefit analysis at all (73%) while nearly 12% report doing a rough estimate, 6% estimate costs only, 7% estimate tangible benefits and costs only, and the remaining estimate costs, tangible and intangible benefits. This result is consistent with other research studies on SMEs. Apparently, either SMEs do not give much importance to the financial consequences associated with implementing new technologies or a majority view the need for such technologies as EDI to be a foregone conclusion. The latter conclusion is also validated by the lack of influence of economic factors on the EDI adoption decision and the great importance attached to customer demands with regards to this decision. As illustrated in Table 8, the nature of cost/benefit analysis conducted by organizations has a significant influence on the ability of a firm to obtain operational/tactical EDI benefits and " a Correlation is significant at the 0.05 level (2-tailed); aa Correlation is significant at the 0.01 level (2-tailed); "'Correlation is significant at the 0.10 level (2-tailed). "This variable has a complex relationship (i.e., it is not linear) with the 'EDI benefits'actors and in consequence the correlation coefficient shown in the table are the nonparametric Kendall's Tau statistic. 55 Journal ofSmall Business Strate/0 Vo/. /3, No. I Spring/Summer 2002 does not correlate with strategic EPI benefits (Benefit2). This result is partly consistent with past research findings reported by Pfeiffer (1992). Volume of EDI communications Greater the volume of messages (documents) exchanged with EDI technology more likely it is for a firm to achieve substantive savings from EDI implementation. The data shows that "volume" is significantly correlated with firms achieving strategic benefits and has no relationship with operational/strategic benefits. Current Stage of EDI Integration ED/ integration is the process during which a firm alters its business practices and applications so that they interface with its EDI application. In this regard, the level of internal integration reflects the variety of applications interconnected with EDI, such as order- entry/purchasing, accounting, production scheduling (MRP), shipping, etc. Another way of defining the level of internal integration is to describe it in terms of stages of integration. Swatman and Swatman (1991),Swatman et al. (1994) have constructed a four-stage model for EDI integration and validated it for large firms. At the lowest level of integration (coded as "I") firms use EDI to print out messages and documents and then re-key data into internal systems. On the other hand, at the highest level of integration (coded as "4") firms use EDI as a strategic technology that links systems throughout the value chain. Obviously, the greater the level of EDI integration the better the opportunity to obtain long-term benefits from this type of technology. The study results (refer Table 8) confirm that the stage of EDI integration is positively correlated with the strategic benefits (ability to compete or forge relationships with trading partners) obtained by sampled-firms. Extent of Trading Partner Support Many authors advocate the use of incentives and subsidies to entice smaller firms to begin using EDI and to expand its use further. This advice has not been always heeded. The trading partners of Kentucky small firms have not heeded this advice as well. Respondents from the sample firms were asked to rate the level of support received from their trading partners for hardware, software, education gt training, telecommunication costs, maintenance and implementation. Respondents used a 3-point Likert-type interval scale to rate each of these categories, with "I" indicating that "no support was received" and "3" indicating that "substantial support was received." Sampled-organizations reported receiving moderate to no support from trading partners in all support categories. The average "support received " score is the highest for EDI implementation (1.76) and Education and Training (1.64). This is consistent with past research and with the fact that "hub" trading partners tend to provide some education/training support and also do pilot testing of new EDI transactions. Intuitively it can be argued that greater trading partner support would translate into higher benefits of EDI implementation for firms. The study data indicates that this assertion is only partly correct. Apparently, at the 95'/0 level of confidence, trading partner support for EDI "software" is the only variable that had a significant positive relationship on operational/tactical EDI benefits (Benefltl) achieved by sampled-firms. Other variables such as trading partner support for "hardware" and "education /k training" significantly influence operational/tactical benefits at the 90'lo level of confidence. All other "support" variables did not have any significant correlation with two 'EDI benefits'actors (refer Table 8). 56 Journal ofSmall Business Strategy Vol. 13, No. l Spring/Summer 2002 Perceived Reasons for Adopting EDI or EDI Decision Criterion Another set of variables that could determine the accrual of EDI benefits relate to the perceived reasons why firms adopt EDI (Iacovou et al., 1995). Two key reasons often touted as highly influential factors for the adoption of EDI in firms are as follows: (Influence of) Customer or Supplier's demand and competitive environment. The correlation results illustrated in Table 9 indicate a different story. On the average, the influence of customer or supplier's demand has no significant relationship with achieving EDI benefits. On the other hand competitive pressures (remaining competitive, pressure from competitors, meeting industry standards) are significantly related to the accrual of strategic benefits in small firms. Table 9: Relationship of 'EDI Adoption Criterion'ith OperationaUTactical Benefits Factor (Benefitl) and Strategic Benefits Factor (Benefit2)' EDI Decision Criterion" BENEFIT I BENEFIT 2 Customer or supplier's demand -.043 (.707) .067 (.557) Remain competitive .131 (.252) .435*'.000) Pressure from competitors .100(.383) .287v (.011) Meeting industry standards .074 (.522) .420' (.000) Improves customer service .440'v (.000) .314 "(.005) Makes Just-In-Time manufacturing possible .203'vv(.075) .079 (.494) Forges strong business relationships with partners .023 (.841) 326vv ( 004) Increases sales revenues/Increases profits .414*v (.000) .244v (.032) Decreases transaction costs .527vv (.000) 263* ( 020) Decreases administrative costs .540v* (.000) .260v(.022) Decreases manufacturing costs .340v'.002) -.015 (.897) Decreases procurement costs .458*v (.000) .075 (.512) Reduces number of employees .455vv (.000) .065 (.576) Reduces inventory & carrying costs .507v* (.000) .042 (.715) Quicker response and access to information .373**(.001) .337"v(.003) Improves accuracy of information .33tvv (.003) .372" (.001) Improves communication with trading partners .186vv (.102) .465v v ( 000) Improves ability to control & coordinate data .369v v (.001) .379vv(.001) Reduces paperwork .357v* (.001) .200 (.079) Ease of processing for order entry .473vv (.000) .224'.049) Aids in accounting, billing, production scheduling 335vv ( 003) .229v (.043) Ease of tracking shipments/Ease of tracking orders .441vv (.000) .217 (.056) Improves efficiency of business operations .50lv* (.000) .155 (.176) Pearson correlation coefficients with significance levels in parenthesis are shown. The useful sample size varies between 77-78 depending on a specific item with the majority of the items having an N of 78." 'orrelation is significant at the 0.05 level (2-tailed; 95% confidence); vv Correlation is significant at the 0.01 level (2-tailed; 99% confidence)." Respondents were asked to assess EDI decision criterion on a 4 point Likett-type scale with verbal labels. Respondents indicated with a check whether a criterion had "no influence at all" (coded as a I), "minor influence" (coded as a 2), "moderate influence" (coded as 3), and "major influence" (coded as 4). 57 Joirrna/ ofSmall Business Strategy Vol. 13, No. I Spring/Summer 2002 Table 10: Relationship of Key 'EDI Implementation Impediments'ith Operational/ Tactical Bene(its Factor (Benefitl) and Strategic Benefits Factor (Beneftt2)"s'DI Implementation Impediments BENEFITI BE/VEFIT? Low volume or frequency of orders -.144 (.212) -.189 (.100) Impersonal nature of EDI -.111(.342) -.153 (.187) Maintaining one system for EDI capable /k another for non- FDI capable panners Translating customer/supplier data for direct use in internal applications Complexity of tbc technology -.177 (.122) -.156(.173) Selecting means for communications with trading partners -.054 (.640) -.041 (.724) Determining appropriate internal applications to apply EDI -.189(.100) -.223» (.051) Ability to seamlessly integrate EDI with existing internal applications Absence of uniform EDI standards .068 (.556) .219(.054) Implementing multiple trading partners -.138 (.231) -.016 (.888) Integrating multiple EDI systems and/or VAN connections -.162 (.158) .221s (.053) Dealing with multiple EDI formats .032 (.781) .273'.015) Selecting the hardware to run EDI software -.253s (.025) -.024 (.832) Changing business processes -.291as (.010) -.032 (.782) Small size of business -.313s'.006) -.190 (.101) Increased responsibility for employees -.152 (.187) -.017 (.884) Oaining management/stakeholder commitment -.024 (.835) -.178 (.121) Overcoming resistance to change -.035 (.765) .010(.929) Availability of managerial time to expand EDI use -.155 (.180) -.063 (.587) Addressing legal issues (e.g., electronic orders, signatures, legal agreements) Exposure to ever-changing customer/supplier requirements about EDI system Managing data and transmission security and auditability .155 (.178) .097 (.401) I ligh startup costs -.078 (.499) .144 (.209) Availability of financial resources -.092 (.425) .106 (.355) High cost of integration and expansion of EDI use -.094 (.412) .171 (.135) Availability of technological resources -.123 (.282) -.068 (.554) Learning new technology and methodology -.181 (.114) -.076 (.506) End users and customers'ontinued reliance on paper-based transaction Obtaining general information about EDI -.226s (.047) -.222'.051) Considering EDI as a natural extension of pre-existing internal operations Understanding potential benefits of EDI -.211 (.066) -.181 (.116) " Pearson correlation coefficients with significance levels in parenthesis are shown. The useful sample size varies between 77 and 78 depending on a specific item with the majority of the items having an N of 78."'orrelation is significant at the 0.05 level (2-tailed; 95% confidence); 'e Correlation is significant at the 0.01 level (2-tailed; 99% confidence)." This is measured on a 3 point Likert-type "seriousness of challenge" scale with verbal labels. A rating of "I" indicates that an item is "not serious at all", "2" indicates that an item is a "somewhat serious challenge", "3" indicates that an item is an "extremely serious challenge." Respondents have the option of indicating that an item is "not an impediment for us" coded as a "0". 58 Journal of Small Business Strategy Vo/. /3, No. l Spring/Summer 2002 Impediments to EDI Adoption and Integration The greater the seriousness and challenge of various impediments to EDI adoption and integration, the lower the chances of increasing or improving the level of benefits afler EDI implementation or integration. Table 10 illustrates the correlation between most common impediments to EDI adoption and integration and the two EDI benefits factors, Although the individual SME owners have told this author that having the "right" volume or frequency of orders is an important challenge, the data in this study indicates that on the average there is no significant relationship between low volume or frequency of orders and the EDI benefits. In fact, most of the more critical challenges that negatively impact EDI benefits have to do with the business process reengineering (BPR) aspect of the technology and the difficulties associated with understanding, modifying or customizing EDI for the adopting firm. Particularly, the difficulty of "selecting the hardware to run EDI soflware", "changing business processes", "small size of business", "obtaining general information about EDI", "understanding potential benefits of EDI", and "considering EDI as a natural extension of pre- existing internal operations" have a significant negative influence on obtaining operational/tactical (direct) EDI benefits. CONCLUDING REMARKS Limitations of the Study As with most research endeavors, this project has some potential limitations. Since the research method used for this study is nonexperimental" in nature, study results are not necessarily generalizable to all SMEs. However, results could be generalized to the industries and organizational sizes represented by the sample. Further, no cause and effect conclusions have been drawn; results are useful for deriving conclusions about relationships and characteristics of EDI use in Kentucky SMEs and similar firms in the larger context. Even though all efforts were taken to reduce nonresponse bias and other errors, inferences, conclusions, recommendations from this type of research strategy are generally supported with lesser confidence than true experimental research (Sproull, 1988). Implications for Practice and Research The results reported in this paper have critical implications for both practice and future research. As suggested in the introduction of this paper, notwithstanding technological developments such as eXtensible Markup Language (XML) and web-based ordering systems, EDI will continue to be a major technological standard for conducting B2B or business-to- business electronic commerce around the globe. The results reported in this study provide some useful guidance for small firms to truly realize benefits in the short- and long-tenn from investments in organization-transforming information technologies such as EDI. Thus, for example, this study demonstrates that regardless of firm size, it is possible to obtain strategic benefits from implementing newer information technologies (IT) such as EDI and that they will not occur in the immediate term (refer Figure I). Further, firms need to give critical consideration to the level of internal integration of the IT being implemented, which has a strong bearing on accruing strategic benefits. In addition, in order to achieve operational/tactical benefits from IT implementation, firms need to better prepare for and understand how they can overcome impediments relating to modifying business processes and choosing the technology itself. 'n experimental variable (e.g., EDI use or non-use) is neither introduced nor controlled in non-experimental research designs. 59 Journal vf Small Business Stra/egv Vo/. 13, //o. l Spring/Summer 2002 Figure I: Signilicant Findings —Determinants of Relative Impact of IT on SMEs Implications for new Interoganizational IT Implementation" SME Characteristics Previous Experience with IT implementation Industrial Sector Volume of Messages IT Benefits Accrued (Documents) Exchanged Strategic Benefits Current Stage of Internal Integration of IT Perceived reasons for IT Operational/tactical benefits adoption: Competitive Pressures Nature of Cost/Benefit Analysis Conducted prior to adopting IT r Trading Partner Support Impediments to IT adoption/integration: Selecting the hardware to run soltware Changing business processes Small size of business Obtaining general information about IT Understanding potential benefits of IT O'onsidering IT(EDI) as a natural extension of pre-existing internal operations " All relationships shown are positive unless otherwise specified. 60 Journal of Small Business Strategy Vol. /3, iyo. I Spring/Summer 2002 Finally, the results of this study provide mixed support for earlier findings by researchers on EDI implementation in small and large firms. The results of this study also show that there are some important determinants and inhibitors of strategic benefits that can be realized by SMEs. Thus, as illustrated in Figure I, significant variables such as the nature of cost/benefit analysis conducted, extent of trading partner support, IT adoption criterion, impediments to adoption/integration, and stage of internal integration and their relationship to strategic benefits accrued from IT implementation in general, and business-to-business commerce technologies in particular, are of clearly of interest to researchers and practitioners alike and warrant further investigation. Acknowledgments: The results reported in this paper were presented at the Kentucky Economics Association (KEA) meeting in Louisville, KY, /999. Tlie author wishes to acknowledge the Kentucky Cabinet for Economic Development for providing partial financial suppor%r this research proyect. REFERENCES Arunachalam, V. (1995, March/April). EDI: An analysis of adoption, uses, benefits and barriers. Journal ofSystems Management, 46 (2), 60-64. Banerjee, S., &. Golhar, D. Y. (1993). EDI Implementation in JIT and Non-JIT Manufacturing Firms: A Comparative Study. International Journal of Operations & Production Management, l3 (3), 25-37. Benjamin, R. I., de Long, D. W., & Michael S. S. M. (1990). Electronic data interchange: How much competitive advantage? Long Range Planning, 23 (I), 29-40. Carter, J. R., Monczka, R. M., Clauson, K. S., & Zelinski, T. P. (1987). Education and training for successful EDI implementation. Journal of Purchasing and Materials Management, (Summer), 13-20. Cash, J. I., &. Konsynski, B. R. (1985). IS redraws competitive boundaries. Harvard Business Review, (March-April), 134-142. EDI IVorld Institute. (1995). Yes, small and medium-size enterprises can do EDI...-and profitably! Retrieved September 4, 1997 from http: //www.ecworld.org/Resource Center/EDIWindow/vol I-no4/feature. html. lacovou, C. L., Benbasat, I., & Dexter, A. S. (1995). Electronic data interchange and small organizations: Adoption and impact of technology. MIS Quarterly, /9 (4), 465-483. Johnston, R. H., & Vitale, M. R. (1988). Creating competitive advantage with interorganizational information systems. MIS (Juarterly, /0 (2), 153-165. Khazanchi, D. (1995, November 22-25). Spoke enterprises: A preliminary assessment of expectations and performance of EDI. Proceedings of the 25th Annual Decision Sciences Institute Meeting, Boston, MA, 880-882. Malone, T. W., Yates, J., & Benjamin, R. I. (1987). Electronic markets and electronic hierarchies. Communications of the ACM, 30 (6), (June), 484-497. McDaniel, C., & Gates, R. (1993). Contemporary marketing research. Minnesota/St. Paul, MN: West Publishing Company. Monczka, R. M., & Carter, J. R. (1988). Implementing electronic data interchange. Journal of Purchasing and Materials Management, 25 (Summer), 26-33. Mukhopadhyay, T., Kekre, S., & Kalathur, S. (1995). Business value of information technology: A study of electronic data interchange. MIS Quarterly, l9 (2), 137-156. Nunnally, J. C. (1978). Psychometric Theory, (2nd Edition). New York: McGraw-Hill Book Company. Pfeiffer, H. K. C. (1992). The dtjtt'aston of electronic data interchange. Heidelberg, Germany: Physics-Verlag. 61 Journal of Small Biisi ness Strategy Vol. I3, No. I Spring/Summer 2002 Phillips 1997 EDI Yellow Pages. (1997). Potomac, Maryland: Phillips Business Information, Inc. Picket, G. C., & Udo, G. J. (1994, March/April). EDI conversion mandate: The big problem for small businesses. Industrial Management, 36 (2), 609. Porter, M. E. (1985). Technology and Competitive Advantage. The Journal of Business Strategy, (Winter), 5(3), 60-78. Porter, M. E., & Millar, V. E. (1985). Flow information gives you competitive advantage. Harvard Business Review, (July/August), 149-160. Premkumar, G., Ramamurthy, K., & Nilakanta, S. (1994). Implementation of electronic data interchange; An innovation diffusion perspective. Journal of Management Information Systems, I I (2), 157-186. Raymond, L., 8c Bergeron, F. (1996). The impact of electronic data interchange on small- and medium-sized enterprises. Journal of Organizational Computing and Electronic Commerce, 6 (2), 161-172. Sproull, N. L. (1988). Handbook ofresearch methods: A guidefor practitioners andstudents in the social sciences, Metuchen, New Jersey: The Scarecrow Press, Inc. Swatman, P. M. C., & Swatman, P. A. (1991, December 16-18). Integrating EDI into the organization's systems: A model of the stages of integration. Proceedings of ihe I2u International Conference on Information Systems, 141-152. Swatman, P. M. C., Swatman, P. A., & Fowler, D. C. (1994). A model of EDI integration & strategic business reengineering. Journal of Strategic Information Systems, 3 (I), 41- 60. Vollmer, K. (2001, April 9). The growth of the internet means business needs more EDI, not less. Internet iyeek, 856, 25. Wilson, T. (2000, February 21). EDI is alive and kicking, study says. Internet Week, 80I, 25. Zorfass, P., & Michel, C. (1992). Electronic data interchange international data corporation white paper supplement to Computerworld. Computerworld, 6(37), SI-S10. Deepak Khazanchi is Professor and Chairperson of the Department of information Sysrems and Quantitative Analysis in The Peter Kiewit Institute (PKI) —College of Information Science &i Technology (IS& T) at the University of Nebraska in Omaha (UNO). He has given seminars and advised firms on seuing up IS'T project management best practices and previously worked as a Project/Design Engineer in the construction industry. His current research interests include philosophical and pedagogical issues in IS, B2B EC control & assurance, and impact ofpervasive IT on service quality. Deepak's recent research has been published in the Communications of the Association of information Systems (CAIS), Journal of the Association of Information Sysiems (JAIS), Decision Support Systems, Information Systems Management, Journal of Information Technology Management, and DATA BASEfor Advances in inform ati on Systems. 62