ELECTRONIC DATA INTERCHANGE: ORGANISATIONAL OPPORTUNITY, NOT TECHNICAL PROBLEM

Paula M.C. Swatman School of Mathematical and Physical Sciences Murdoch University South Street, Murdoch, W.A. 6150. Paul A. Swatman School of Computing Science Curtin University of Technology G.P.O. Box 1982 Perth, W.A. 6001

ABSTRACT Electronic Data Interchange (EDI) enables organisations to send and receive standardised business communications more quickly, flexibly, cheaply and with greater security and accuracy than is possible with conventional postal services. It is not, however, merely another telecommunications advance, but rather provides a means by which organisations can introduce seamless inter-connection within and across organisational boundaries. This paper summarises the background and development of EDI and the benefits to be obtained from its integration into internal and external organisational systems and considers the future of EDI and inter-organisational information systems in general. It then discuss the organisational issues involved in implementing EDI, dispelling the myth that EDI is an issue relevant only to computer communications professionals, pointing out the fact that EDI is a major strategic opportunity which must be addressed at senior levels within implementing organisations.

This paper was presented to "DBIS ’91" the 2nd Australian Conference on Database and Information Systems, held at the University of New South Wales in February 1991. It has since been republished in Srinivasan B. & Zeleznikow J. (Eds.) Databases in the 1990’s: 2, World Scientific Press, Singapore, 354-374.

INTRODUCTION The many, similar definitions of Electronic Data Interchange (EDI) are succinctly summarised in the following quotation from Emmelhainz (1988): "computer-tocomputer transmission of standard business data". A Datamation survey of EDI usage points out that the original attempt to standardise documentation in the U.S. transport industry during the late 1960’s has now developed into a multi-million (or even multi-billion) dollar industry: "by 1992, 80% of European retailers will be conducting business via EDI links" (Lamb 1988). Despite the growing body of literature and industrial interest in EDI, there is still debate concerning the topic’s definition. A detailed description of what EDI is (and is not) is spelt out in a paper by Brawn (1989), in whose view the essential elements of EDI are: • direct application-to-application (not merely computer-to-computer) communication; • the use of an electronic transmission medium (normally a value-added network) rather than magnetic tapes, disks, or other transmission media; • the use of electronic mail boxes for "store and collect/store and forward" transmission/delivery of documents; • the use of structured, formatted messages based upon internationally agreed standards (thus enabling messages to be translated, interpreted and checked for compliance to a standard set of rules). The term EDI therefore does not refer to: • electronic mail (which must be read by the recipient and which does not make use of standardised document formats); • file transfer (which also makes little use of standardised formats and which need have no connection with applications at either end of the transmission); or • remote data entry (which merely places the entry terminal some distance away from the computer). This standards-based form of communication gains its attractiveness from promises of huge savings in inter-company transaction costs and greatly increased productivity. EDI is already seen as an essential prerequisite for Just-in-Time/Quick Response inventory and production systems (Robinson and Stanton, 1987; Sadhwani and Sarhan, 1987; Knill, 1989; Skagen, 1989), where limited stock holdings are dependent on rapid and accurate communication of purchasing needs. Inter-connections are now being urged between EDI and other types of information systems - such as electronic funds transfer (EFT) systems, where the combined EDI/EFT systems cater for the exchange of both payments and remittance information and are known either as Electronic Payments (EP) or as Financial EDI (Hill and Ferguson, 1987; Canright, 1988; Cafiero, 1989); or corporate application systems (Svinicki, 1988; Wilmot, 1988) to enhance integration. Despite its attractiveness and the wide-spread public interest demonstrated by such things as conferences and articles in the "trade press", EDI growth world-wide has 1

not demonstrated the explosive tendency which was widely predicted during the mid-to-late 1980’s. Skagen (1989) suggests that in the U.S. "steady gains of 1,500 to 2,000 companies a year [on a user base of around 5,000] characterise the growth curve ... but the figures are a far cry from predictions made by some of EDI’s early champions [who] claimed, for example, that 70 percent of all businesses would be making significant use of the technology by 1993". Barriers to EDI Diffusion There are two possible explanations for this slow-down in growth, which to some extent overlap. The first obstacle to the diffusion of EDI is the difficulty of quantifying its benefits, causing organisations which are not convinced that EDI will be of real use to defer the decision to participate. Surveys and case studies undertaken in the U.S. (EDI Research, 1988; Benjamin et al, 1990) indicate that the real benefits available from EDI are qualitative in nature - the direct financial gains from EDI tend to be small and rather unimpressive. Despite the overall gains recorded by organisations such as General Motors, who claim that they have reduced the cost of each car by $200 (Rochester, 1989), major benefits arise indirectly, that is, they lie in the areas of inventory control and improved trading partner relationships. Not all organisations are sufficiently far-sighted to make the sort of investment required to obtain these benefits, nor are they sufficiently perspicacious to realise that these longer-term benefits are attributable to EDI. As an example, Just-in-Time inventory management receives the plaudits which would not be possible without EDI’s standardised communications infrastructure. The second impediment relates to EDI’s image as a technical problem - many organisations ignore the fact that without a unified plan and management support EDI schemes will tend to become "perpetual pilots", going nowhere and achieving little of real benefit to their organisation (Gannon, 1990; Swatman and Everett, 1990; Swatman and Fowler, 1990). Even in those organisations aware of EDI, the concept is often seen as an issue for the more technical members of the I.T. Department (see Fowler and Swatman, 1991). The problem of quantifying EDI’s potential benefits has been discussed in a number of recent papers (see for example Bedford, 1988; Skagen, 1989; Rochester, 1989). This paper focuses on the second obstacle to the wide-spread implementation and diffusion of EDI, its "technical" image, pointing to the need for organisations to take a more strategic view of the issue - that is, for senior management involvement in planning for EDI systems. The paper initially places EDI technology in its overall context. It then discusses the limited technical concerns of EDI implementation, such as standards (both those relating to the data communications infrastructure supporting EDI and those relating to the conversion of documents between the various existing in-house, intra/interindustry, and international formats); standardisation of such items as codes and identifiers amongst the members of industry groups; and system development issues, particularly those pertaining to the design/implementation of EDI systems. We then consider the organisational issues and implications of EDI, such as: integration with 2

in-house application systems; organisational readiness for EDI; change management concerns; legal and security implications; and senior management support in planning and overall coordination, pointing to their crucial role in EDI’s long-term success. Finally, the relative importance of the technical aspects of EDI is compared with that of the wider organisational, managerial and strategic issues involved. EDI IN CONTEXT Historical Development of EDI The genesis of EDI can be traced back to the 1948 Berlin Airlift, where the task of coordinating airfreighted consignments of food and consumables was addressed by devising a standard manifest to be filled in by aircraft before unloading (Brawn, 1989). It was not, however, until the 1960s that the rail and road transport industries first began to standardise documents and replace paper-based communication methods, with the formation of the US Transportation Data Coordinating Committee to oversee the development of standards for all groups of carriers (McNurlin, 1987). These standards, however, were not truly universal - being an intermediate translation stage between four existing sets of industry-specific standards. A more significant move towards standardisation came with the ANSI X12 standards, which were developed in the U.S. at about the same time that the United Nations Economic Commission for Europe (UNECE), under pressure from the U.K. Customs and Excise, was developing its own set of standards for documents used in international trade (Schatz, 1988). The U.K. standards became known as GTDI (General-purpose Trade Data Interchange standards) and were gradually accepted by some 2,000 British exporting organisations. This otherwise useful development of standards began to cause problems on its own account early in the 1980’s, with North American and European organisations using quite different (and incompatible) sets of standardised documents. A major step towards a solution for this problem was taken with the formation of a joint European and North American working party (Carter, 1989), which is currently developing the EDIFACT (Electronic Data Interchange for Administration, Commerce and Transport) standards, now recognised internationally (ISO 9735). At present only two EDIFACT documents are complete (an invoice and a purchase order), but a full range of business documents is planned for release over the next few years. Competitive Weapon or Cooperative Platform Industry conferences and journal articles frequently describe EDI as providing a "competitive edge" for its users (e.g. Robinson and Stanton, 1987; Langton, 1988; Hudson, 1989). While short-term competitive advantage from EDI is possible for an organisation which is the first in its market sector to link in suppliers, the arrival of competitors rapidly erodes any possibility of "sustainable" competitive advantage (Clemons, 1986). Some instances of sustainable competitive advantage through EDI do exist, but these appear to be more likely where: 3

• • •

EDI is a factor in a substantial gain in market share being made and retained, particularly if some competitors are forced to withdraw from the market (e.g. the U.S. airline industry’s reservation systems SABRE and APOLLO); alliances are formed which provide dominance over a market (e.g. the international banking community’s closed-user-group SWIFT clearing-house, or the European airline reservation clearing-house SITA); and/or the market is substantially re-defined in the initiating organisation’s favour (e.g. the American Hospital Supply Company’s on-line ordering system).

It is noticeable that, with the exceptions of the SWIFT and SITA clearing-houses (which clearly fall within the cited definition of EDI) most of the examples of sustainable competitive advantage of this type relate to systems which many observers would consider only border-line EDI at best. In most cases of this type, the innovator’s advantage is fairly quickly neutralised, either because the original competitive scheme is replaced by a more successful cooperative venture, or because the innovators are forced by competitors or regulatory agencies to make the services widely available. The successful implementation of cooperative EDI schemes within such fiercely competitive market sectors as the automotive and pharmaceutical industries implies that corporations view EDI as a factor in their industries’ survival, rather than as a competitive weapon. In fact, the competitive SWIFT venture is itself an example of cooperation by the banking industry, to protect its market share against the rapidly expanding non-bank financial institution sector. This failure of EDI to provide long-term sustainable competitive advantage implies that, despite its strategic importance, EDI is essentially a cooperative rather than a competitive phenomenon: "Implicit in the early writings on EDI has been the assumption that these systems hold great potential for providing strategic advantage ... [but] EDI applications, rather than being a competitive weapon, are increasingly a necessary way of doing business" (Benjamin et al, 1990. See also McNurlin, 1987; Rochester, 1989; Swatman and Clarke, 1990). There are competitive advantages to be obtained from the use of EDI, but these relate to the integration of EDI into the organisation’s internal structure and systems - and are therefore more appropriately included in the section of this paper concerned with the organisational implications of EDI. Benefits of EDI EDI’s direct impacts include labour-savings in the areas of data transcription, controls, and error investigation and correction; and fewer delays in data-handling. The lists of the resultant benefits which are to be found in the business literature are remarkably similar, indicating that organisational experiences have been fairly consistent. A typical list of benefits is provided by Emmelhainz (1989): • improved internal operations from a reduction in time; • better responsiveness to customers;

4

• •

improved trading partner relationships; increased ability to compete, both domestically and internationally.

As organisations gain greater experience with EDI, the more far-sighted users have realised that significant benefits are to be gained through the integration of related functions across organisational boundaries. It is not the replacement of electronic messaging for formerly paper-based communications which provides EDI’s major strategic capabilities, but the associated changes in operation and function within and between organisations, made possible by EDI links. For example, Rochester (1989) cites the case of Levi Strauss which, through its Quick Response system LeviLink, has achieved a complete vertical integration of the company’s entire apparel manufacturing and marketing cycle (including the replenishment of inventory, the management and reconciliation of purchase orders, the receipt of goods, the processing and payment of invoices, the capture of point-ofsale information and the analysis of market trends). The management of Levi Strauss sees EDI as a unifying factor in the organisation’s use of Quick Response technology (DuBois, 1990), providing the company with a significant edge in a highly competitive market-place. This focus on achieving integration across organisational functions and between organisations is what distinguishes EDI from other forms of electronic transaction - and makes EDI an effective strategic application at organisational, national and international levels. EDI AS A TECHNICAL ISSUE The Standardisation Issue A major issue in systems development is the growth of standards of all types: from data communications to applications software interfaces. EDI systems have provided further impetus to the development of international standards at two separate levels: not only is EDI responsible for the rise of document translation standards, with the introduction of ANSI X12 and EDIFACT; but it has added pressure to calls for commonality of data communication standards. The need for standards at both these levels is widely appreciated (Scott, 1988; Carter, 1989; Hollands, 1989) although EDI is more commonly associated with the development of document translation standards. The opportunities for additional value-added services from the integration of EDI and X.400 messaging, however (such as delivery and non-delivery notification, time stamping, multi-destination delivery, grade of delivery selection, deferred delivery and hold-for-delivery) adds weight to the case for truly international data communications standards - and thus to the claims of Open Systems Interconnect (OSI). "One of the factors most clearly distinguishing EDI from other [inter-organisational systems] is the open standardisation of document formats, which enables cross-connection between systems and allows users of one EDI system to join other, similar systems with comparative ease." (Swatman and Swatman, 1989).

5

This movement is being accelerated by the formation of Europe’s Single Unified Market (the EC’92 programme): "In fact, the greater standardisation of systems - both those from suppliers and those being used within user organisations - is likely to be one of the most significant consequences of the single market" (Tate, 1989). It appears that EDI, rather than creating technical problems of standardisation, has instead assisted in finding solutions to these problems. One aspect of standardisation which does raise a practical concern amongst those implementing EDI systems is the need for a consistent set of codes and identifiers between the various system users. Wilkinson (1989) discusses the problem in the context of an Australian EDI system designed for tracking railway containers, where each State railway authority uses different codes to designate the same items. The Railways of Australia scheme makes use of a simple conversion table to solve this apparently complex problem. It may well prove that the major obstacles to agreement on inter-company nomenclature are organisational or political in nature, rather than technical. EDI System Types - a Development Perspective The view that EDI is a technical problem which must be solved anew for each system is reflected in the approach often taken to implementing new EDI schemes (Swatman and Fowler, 1990) - that of handing the problem to the IT Department to solve. EDI software, in fact, falls into two quite separate categories - the in-house software which translates information from unstructured, company-specific formats into structured EDI formats (such as ANSI X12 or EDIFACT); and the communications software which transmits the structured message to its recipient. The transmission process can take place in one of two ways - either by linking the two partners directly (point-to-point) via modem, or by means of third-party networks: "in essence, what a third-party network provides is the EDI communication skills, expertise and equipment necessary to communicate electronically. In addition, [it] can also provide value added services such as translation to standard, international connections, connection to other third party networks and training" (Emmelhainz, 1989). Organisations contemplating the implementation of EDI tend to see each scheme as unique, largely because each company involved in a particular market segment transacts its business slightly differently from its competitors. It was this perception that a company’s methods of doing business were unique (and thus software to meet those needs must also be unique) which led to the development of so many software variations on universal requirements such as payroll or general ledger. In much the same way, despite the perception of EDI as an ever-new technical problem, EDI systems themselves generally belong to one of three categories (Akerman and Cafiero, 1985): 1. one-to-many systems: These systems typically arise when a (large) organisation wishes to streamline its interactions with suppliers (or customers). The initiating organisation is the hub of the system, while its trading partners form the satellites. 6

> < > >

Figure 1(a) - One-to-Many EDI System 2. many-to-many or "clearing house" systems: A more general form in which there is no single hub but, apparently, many buyers and sellers interacting with each other. Notionally, the system itself forms the hub and all parties are satellites. The development of a system of this type is usually driven by two organisations, each representing an industry group. In a sense, these systems can be considered as one-to-one systems connecting the buyer industry group with the supplier industry group:

< < < > <

> > >

Figure 1(b) - Many-to-Many EDI System Akerman and Cafiero do not carry this idea further, although it is an easy task to extend this concept to allow for the participation of any number of industry groups - each of which may be either supplier or customer.

< < < < > > < <

> > > >

Figure 1(c) - Generalised Many-to-Many EDI System This extension caters for such complex inter-sectoral groupings as the U.S. automotive industry scheme (McNurlin, 1987), where the powerful chemical industry can be seen as a sector in its own right (although it plays a comparatively small part in the creation of an automobile). The generalised 7

model is sufficiently complex to allow for wider extensions to the concept of Inter-Enterprise Systems (Ellinger, 1989), such as manufacturing organisations’ Just in Time inventory management schemes, or the wholesale/retail industry’s Quick Response systems. 3. "incremental paper trail" systems, where documents are amended by a series of participants with additional information being added to the document at each stage in the process, are particularly relevant to the domestic and international trading community:

> > > > V V V V

Figure 1(d) - Incremental Paper Trail EDI System Schemes of this type allow the progress of a shipment from exporter to importer to be covered by a single electronic document, to which each party merely adds the appropriate information, obviating the need for bulky and unmanageable documents such as bills of lading or ships’ manifests. Despite the apparently diverse appearance of the system sub-types, such representations of EDI systems can be replaced by a generalised EDI system structure. Winston (1984) suggests that when facing a new problem, time is better spent in defining an appropriate representation of the problem (and its inherent constraints) than in devising clever inference and control mechanisms. A systems view of EDI schemes would enable EDI users to see each new implementation as a variation on a theme, rather than as a completely new technical problem. It is clear that Figure 1(c) represents a generalisation of many-to-many EDI systems; and that many-to-many systems are themselves a generalisation of one-to-many systems. The single differentiating characteristic of "incremental paper trails" is the existence of persistent memory within the EDI system itself. Figure 2 (below) provides an illustration of a generalised EDI system, incorporating the concept of persistent memory, necessary for incremental paper trail documents. Although this diagram shows a single data store for the entire EDI system, we do not imply that the data store could not be implemented as a distributed database. In fact, 8

the current European development of an electronic Bill of Lading (EDI Monthly Report, September 1990) with its consequent need for multiple third-party network providers, seems to indicate that distributed document storage will be the method of choice (if not actually that of necessity). It is now clear that we can redefine EDI at a systems level in the following manner: EDI systems are that sub-class of distributed information systems where the users of the system are organisations rather than individuals.

> > < < >

<

> > < EDI

< > Data Store System <

Figure 2 - A Generic Model of an EDI System The rapid growth of third-party EDI network service providers (Takac and Swatman, 1989) offers further support for the proposition that the provision of EDI services is, at worst, a solvable and comparatively standard technical issue. From the systems developer’s perspective, therefore, it can be said that EDI is not a new problem, although such a statement does not imply that the development of EDI systems is a simple issue. The successful development of large, organisationally effective systems continues to be an area in which we achieve only limited success. The most significant consequence of defining EDI in terms of a single system type is not that it removes the difficulty from implementing EDI schemes, but that it debunks the myth of EDI’s technical complexity. Clearly, from a technical point of view, the fact that the users of a system are organisations (rather than individuals) is largely irrelevant. EDI per se offers few unique technical problems - despite its organisational challenges. EDI AS A STRATEGIC ORGANISATIONAL ISSUE Although EDI itself is simply a standards-based approach to transmitting information, the strategic ramifications of its co-operative focus are significant. As we have already mentioned, in discussing the cooperative gains available from EDI, longerterm benefits are obtained from the integration of intra and inter-organisational functions. No matter how one approaches EDI, from the organisational viewpoint 9

all approaches lead back to the need for a consistent, centralised and strategic view of this enabling technology. Organisational Readiness for EDI Large organisations, with well-developed information systems facilities, may have little difficulty in coming to terms with the integration problems inherent in EDI. The average small to medium organisation, however, may not have gained expertise in information system integration or, more generally, have acquired a corporate view of information systems. Users of external services such as corporate cash management systems, for instance, rarely see the data they receive as an integral part of overall information management. Organisations’ views on (and approach to) EDI can vary widely depending on whether they were the original instigators of an EDI scheme (a proactive focus), or were persuaded to join an existing system by a major customer (a reactive focus). The proactive organisation is able to control the implementation of EDI from the top down, designing systems and choosing partners in the most appropriate manner. The reactive organisation, by contrast, may well have had no desire to become involved with EDI. Trading partner pressure has persuaded many small suppliers to connect to an EDI scheme whose implications and long-term benefits are imperfectly understood. Fowler and Swatman (1991) reports a survey of Western Australian exporters’ interest in EDI and the influence of the EC’92 programme; and finds the trading partner argument a compelling motive for EDI implementation. A strategic, organisational view of EDI is clearly more likely to lead to a proactive approach to the setting up of EDI schemes, with the benefits which accompany control. It should, however, be noted that even those smaller organisations whose participation in EDI schemes is undertaken as a result of pressure from powerful trading partners, can benefit from this forcible involvement: "it has often been suggested that the benefits of EDI are greater for larger companies. In reality, those same benefits can be far more valuable to smaller, more budget-conscious organisations who incur many of the same expenses but don’t have the multiple areas of revenue to offset the cost of doing business" (Lyttle, 1988). Because EDI involves a direct, almost intimate relationship with trading partners (Emmelhainz, 1988; Payne, 1989; Skagen, 1989), inter-organisational relationships must be established on a firm contractual footing, with questions of accountability and liability defined and clarified. Although the potential for long-term growth on both sides makes the creation of such complex arrangements well worth while, newcomers to the field must be aware that the payback may be a long time in coming. It is clearly a top management decision to undertake long-term investments which may not provide positive cash flow benefits for several years. Further, the manner in which such organisational restructuring is introduced into an organisation will affect both the success of the restructure and its acceptance by management and staff alike.

10

Integration with Existing Application Systems Strategic benefits attributable to EDI depend upon the integration of information received from external sources with existing organisational systems and practices and the integration of those systems and practices themselves - which can change the entire structure of the affected organisation. During our discussion of the competitive/cooperative aspects of EDI, we suggested that EDI is essentially a cooperative activity. It should be noted, however, that while EDI per se is not a competitive weapon (except in the very short term), there is immense scope for competitive advantage to be gained from the way in which EDI in integrated into the organisation’s structure: "I suggest you look particularly hard at intra-company activity, and look beyond purely paperless trading for the early international paybacks ... Installing an inter-company point-to-point EDI link for trading is trivial compared to the thousands of internal information links that are also needed if EDI is to bring optimum benefit. In reality this is why the most successful users of EDI don’t mind joining industry clubs and sharing the benefits of intercompany EDI, because they have usually got massive organisational restructuring and internal information systems plans to exploit the situation competitively" (Wilmot, 1988). Inter-connections of this sort alter an organisation’s day-to-day operations, reducing or eliminating administrative overheads such as overtime premiums, overnight courier charges, late or incorrect shipments from suppliers, excess inventories, poor forecasting, incorrect order entry and disruptive production schedules. Such a major internal restructuring cannot be achieved by the technical experts of an organisation’s I.T. Department, but requires support from the senior executive levels. Sustained competitive advantage is a goal of virtually all private sector corporations (and many public sector organisations) - providing this capability by investment in technological infrastructure is as much a top management decision as are the decisions to diversify product lines or acquire new subsidiaries: "IS guidelines are a logical extension of top management’s traditional role of setting business strategy and objectives. By setting such guidelines, a CEO can bridge the gap between business goals that seem almost unobtainable and the systems that can allow the goals to be met" (Tinsley and Power, 1990). Change Management Considerations To control the introduction of EDI properly, we must phase it into an organisation: "By its very nature, the complete integration of EDI within a company requires the coordination of purchasing, accounting, transportation, systems, legal and auditing personnel. Further, because EDI is intercorporate, the departmental coordination needs to be mirrored in the trading partner’s organisation. All of this makes the EDI change process one that requires careful planning and close management" (Emmelhainz, 1989:125).

11

One of the crucial issues of EDI’s potential for organisational restructure is that of human resources - and particularly in the area of staff retraining. Perry (1989) points out that there are two aspects of EDI’s human resources issue - changes to management function (including the threat of being by-passed, reduced supervisory involvement and the need to deal with staff trained in new and unfamiliar techniques); and changes in staff functions (including award restructuring, job redefinitions, modified job classification structures and redeployment/redundancy requirements). In some senses EDI is no different from any electronic innovation in that it will cause the loss of some jobs and a consequent need to re-train existing and new staff. There is, however, one area in which EDI differs significantly from other electronic business initiatives. The greater efficiency EDI makes possible will have profound implications for working conditions in industries such as those on the waterfront, where conditions have changed little for centuries (consider the response of the British printing unions to the introduction of photo-typesetting to gain a perspective on this issue). Perry (1989) provides evidence that the Australian union movement has already foreseen EDI’s potential to cause massive industry restructuring: "These factors are rarely, if ever, acknowledged in the EDI literature (which tends to focus on technological issues) ... Any suggestion that this is raising an unnecessary spectre can be answered by reference to the [September 1989] In-Principle Agreement negotiated between the ACTU [the Australian Council of Trade Unions], stevedoring employers and stevedoring unions, under the auspices of the Waterfront Industry Reform Authority: The functions of the Authority include the provision to: ’Seek the co-operation of relevant organisations not party to this Agreement for the purpose of ... (ii) introduction of Electronic Data Interchange and revised standardised shipping and waterfront documentation’ [4.2(f)] Under Award/Job Restructuring, the Agreement states: ’New classification structures shall pay regard to the progressive introduction of Electronic Data Interchange in the industry and its overall impact’ [13.9]". Many current union activities involve complex and inefficient practices which have the potential to obviate EDI’s streamlining of business activity. The challenge is to identify and maintain levels of personal expertise in business operations and customer relations, without retaining the manual processes with which they are currently associated - and to achieve this in a manner which will benefit both union members themselves and the organisations which employ them. Negotiating with major union groups in the interests of both the organisation and its employees is clearly a task which requires a coordinated, top level management approach. Senior Management Support As with any major new technology, effective implementation can only occur when there is support from the top down. In a 1988 survey conducted by the U.S. 12

company EDI Research Inc (1988), lack of management commitment was rated fifth in importance of barriers to EDI by large respondents (although smaller respondents did not see this factor as a major obstacle). The need for company-wide support for EDI is stressed in a report by the EDI Council of America (EDIA), where the issue of restructuring a company to provide full integration between an incoming EDI message and order entry, sales tracking, inventory management and manufacturing systems is discussed (quoted in Rochester, 1989): "To accomplish this [integration], the EDI system must be designed to achieve the company’s goals. And for this to occur, top management must support a total EDI strategy. Information systems is in an ideal position to promote such a strategy and implement it throughout the organisation. Information systems can see the links between various departmental applications accounting, engineering, manufacturing and so forth - and then work with the appropriate business managers to determine in what ways EDI would be beneficial" While concurring with the importance of top management support, an alternative view is presented in a Butler Cox report on EDI (Butler Cox, 1987) which suggests that "the organisation needs to prepare itself for taking action on EDI. Major strategic applications usually need to be championed by a senior manager with a role related to the area of application, rather than by the systems department. In large, diversified organisations a complementary activity, based in the systems department, may also be required. The purpose of this activity is to promote EDI in general rather than a particular EDI application". In order to assess the extent to which technical issues of integration have restricted EDI’s growth within Australia, the authors conducted a case study of Australian organisations involved in EDI (Swatman and Swatman, 1990). The results of this study indicate that those technical problems which do exist are solved in almost exactly the same way in all organisations - the major barrier to rapid EDI system growth in Australia appears to be the piecemeal nature of many of the schemes and the lack of a coordinated, top management approach to the topic. These findings show a fair degree of similarity with those of an MIT study undertaken to assess EDI’s ability to provide competitive advantage: "Despite optimistic assumptions about the ability of developers to expand EDI use beyond the prototype stage, significantly increasing the penetration of these systems often proves more difficult than expected. Moving EDI systems into production means confronting not only barriers imposed by a lack of standards, but also the organisation’s resistance to large-scale change. Top management support in this process is essential because experience indicates that the ability to implement a successful EDI pilot, while essential for learning, is no guarantee that the application can be readily expanded to a broader group of users" (Benjamin et al, 1990).

13

EDI, with its offering of a single, standardised interface between otherwise incompatible information systems, is essentially a facilitator of organisational change and will therefore provide its most effective benefits only when implemented in a top-down manner by senior management. A survey of EDI use in within the Western Australian public sector (Swatman and Fowler, 1990) showed that there was considerable concern amongst senior IT staff at the lack of top management support for EDI use. In some cases EDI schemes had been introduced unofficially and in all but one case, there was little or no evidence of a strategic approach to the development and implementation of EDI. CONCLUSION We believe that the conventional views of EDI, which envisage the field as being either a sophisticated tele-communications problem, or a vehicle for short-term competitive advantage, are merely the tip of the iceberg. The major benefits of EDI lie rather in the areas of inter-connectivity and integration of inter- and intraorganisational information systems and communications media. The present trend towards third-party software supply (see Dearden, 1987 and Swatman et al, 1990) means that organisations are increasingly faced with the difficulties associated with exchanging information amongst a multitude of incompatible software packages. EDI, with its standardised document exchange facilities, offers an entirely new way of solving this problem - without the need for a single, monolithic database and system structure. Organisations which view EDI as falling within the domain of the technocrats will fail to obtain the benefits of seamless interconnection between the organisation’s own internal application systems and across organisational boundaries to its trading partners. While there are some technical issues involved in EDI implementation, "technologically, EDI is not complex. Further, the EDI structure of standards, software and networks is in place and is working. The technological implementation issues need to be understood and addressed by firms using EDI; however, these issues do not usually present significant difficulties to most firms" (Emmelhainz, 1989). Emphasis is currently focused on the technical issues of connection to EDI networks; and at the micro level of single EDI networks. That focus should rather be the integration of EDI into the corporate information systems plan. Organisations which persist in viewing EDI as an isolated, technical problem for the IT department or the tele-communications experts to solve will find that their EDI systems fail to expand far beyond their original pilot stage, which has been the Australian experience until now. Competitive advantages from this technology are likely to be short-lived, although those organisations not able to grasp the EDI opportunity will find themselves increasingly isolated from the business community. Co-operation amongst and within industry groups is the key to achievement of the significant benefits offered by EDI. Such cooperation can only be achieved by organisations in which senior management has recognised the importance of EDI - middle-level management does not possess sufficient influence to ensure industry group agreement.

14

As the survey of Western Australian exporters indicated, Australian organisations are tending to wait for pressure from large trading partners before taking any steps in the direction of EDI implementation. Pressure of this sort is already occurring and can only continue to increase, with the imminent completion of the EC’92 programme. Reactive organisations which implement the minimum EDI connection will fail to obtain the real strategic benefits possible for those proactive organisations which view EDI from the top down and implement a consistent, coordinated and seamless approach to information system integration. The most crucial issues, ones which must be addressed as a matter of urgency by Australian organisations, are those of managerial support and involvement in the whole area of standardised system integration. Australia is a small trading nation and cannot afford to stand out against the technological/organisational innovation of EDI which is already being embraced by the major trading blocs such as Europe and North America.

15

REFERENCE LIST Akerman G.C. and Cafiero W.E. (1985) Introduction to Electronic Data Interchange: A Primer, GE Information Services, 1985. Bedford W. (1988) The Practical Issues, Proc IPS Conf "EDI Benefits and Opportunities", London. Beer B. (1984) Legal Aspects of Automatic Trade Data Interchange, Transnational Data Reporter Vol 7, No 1, Jan-Feb, 52-7. Benjamin R.I., De Long D.W. and Scott Morton M.C. (1990) Electronic Data Interchange: How Much Competitive Advantage?, Long Range Planning, Vol. 23, No. 1, 29-40. Brawn D. (1989) EDI Developments Abroad and How they Impact on Australia, Proc IDC Conf "EDI - The Key to Profitability in the 1990’s", Sydney, December. Brown R. (1989) EDI and Proving Your Transaction, Proc Conf "SOST ’89", Terrigal, NSW, 7-16. Butler Cox Foundation (1987) Electronic Data Interchange - Research Report 59, Butler Cox and Partners Limited, London, September. Cafiero W.G. (1989) VANs, Not Banks, Best Suited to Deliver Remittance Data, Corporate Cashflow, April, 34-38. Canright C. (1988) Banks Seek EDI Opportunities, Bank Administration, April,40-41. Carlyle R.E. (1988) Managing IS at Multinational, Datamation, March 1, 54-66. Carter S. (1989) Standards for EDI and Communications, Proc IIR Conf "Gain the Competitive Edge with EDI", Melbourne, May. Clemons E.K. (1986) Information Systems for Sustainable Competitive Advantage, Information and Management, November. Copeland D.G. and McKenney J.L. (1988) Airline Reservations Systems: Lessons from History, MIS Quarterly, 12/3, 353-370. Dearden J. (1987) The Withering Away of the I.S. Organisation Sloan Management Review, Summer, 87-91. DuBois G. (1990) How to Improve the Bottom Line and Reduce the Cost of Doing Business -How Does EDI Fit In? Proc IBC Conf "Paperless Trading - Strategies for the 1990’s", Sydney, March. EDI Research (1988) The State of U.S. EDI: 1988, EDI Research Inc. November.

EDI Research (1988) The State of Canadian EDI: 1988, EDI Research Inc., November. Ellinger R. (1989) An IESF Primer: What it is and Why, Manufacturing Systems, December, 31-33. Emmelhainz M.A. (1988) Electronic Data Interchange: Does it Change the Purchasing Process? Journal of Purchasing and Materials Management, Winter 1988. Emmelhainz M.A. (1989) Electronic Data Interchange: A Total Management Guide, Van Nostrand Reinhold, New York. Fowler D.C. and Swatman P.M.C. (1991) EDI and Europe 1991: I.T. Solutions for Tomorrow’s Exporters? submitted to Proc Conf "SOST ’91", Adelaide, October (copies available on request to the authors at Murdoch University). Gannon P. (1990) How Long Must EDI Wait in the Wings? Computer Weekly, February 15. Hanson J.V. and Hill N.C. (1989) Control and Audit of Electronic data Interchange, MIS Quarterly, Vol 13, No 4, December, 403-412. Hardy M. (1988) Opening up European Telecommunications - the Commission Perspective, Proc Conf "EDI ’88", London, November. Hill C. (1988) Getting Started, Proc Conf IIR "EDI: Profiting from Paperless Transactions", Sydney, October. Hill C.M. (1989) EDI - The Competitive Edge, Proc Conf "SOST ’89", Terrigal, NSW, May, pp.193-208. Hill N.C. and Ferguson D.M. (1987) EDI and Payment Terms: Negotiating a PositiveSum Game, Journal of Cash Management, Sept/Oct. 21-26. Hollands D. (1989) Electronic Data Interchange (EDI) in the Australian Automotive Industry, Proc Conf "SOST ’89", Terrigal, NSW, May. Hudson P. (1989) EDI The Emerging Technology for Document Productivity, Proc Conf "GTE ’89", Canberra, March Knill B. (1989) Warehousing 89: Focus is on Quick Response, Material Handling Engineering, March, 53-60. Lamb J. (1988) IBM Eyes EDI in Europe, Datamation Vol 34,No 13, July, 48.12-48.16 Langton Limited (1988) EDI for Competitive Advantage, Special Paper - Proc Conf "EDI ’88" London, November. Lyttle R.M. (1988) Will You be Proactive or Reactive on EDI?, Systems 3X World, 16, 11, November.

Payne R.A. (1989) EDI Implementation: A Case Study, Journal of Systems Management, March, 14-20. McNurlin B.C. (1987) The Rise of Co-operative Systems, EDP Analyzer, Vol 25, No 6, June, 1-16. Newman M. (1989) Some Fallacies in Information Systems Development, International Journal of Information Management, Vol 9, No 2, 127-143. Norris D.M and Waples E. (1989) Control of Electronic Data Interchange Systems, Journal of Systems Management, March, 21-25. Perry, N. (1989) Change Management in the EDI Environment, Proc IDC Conf "EDI The Key to Profitability in the 1990’s", Sydney, December. Robinson D.G. and Stanton S.A. (1987) Exploit EDI Before EDI Exploits You, Information Strategy, The Executive’s Journal, Spring, 32-35. Rochester, J.B. (1989) The Strategic Value of EDI, I/S Analyzer, Vol. 27, No. 8. Sadhwani A.T. and Sarhan M.H. (1987) Electronic Systems Enhance JIT Operations, Management Accounting, December, 25-30. Schatz W. (1988) EDI: Putting the Muscle in Commerce and Industry, Datamation, Vol 34, No 6, March, 56-64. Scott P. (1988) Trends in Communications, Proc Conf "EDI ’88", London, November. Skagen A.E. (1989) Nurturing Relationships, Enhancing Quality with Electronic Data Interchange, Management Review, February, 28-32. Svinicki J. (1988) Integrating EDI Into the Second Supplier Tier, Systems/3XWorld, September, 70-76. Swatman P.M.C. and Clarke R. (1990) Organisational, Sectoral and International Implications of Electronic Data Interchange, Proc IFIP Conf HCC4 (4th International Conference on Human Choice and Computers), Dublin, July. Swatman P.M.C. and Everett J.E. (1990) Australian Involvement in Electronic Data Interchange - 1989, Australian Computer Journal, forthcoming. Swatman P.M.C. and Fowler D. (1990) Electronic Data Interchange (EDI) Within the State Public Sector, W.A. Department of Computing and Information Technology Position Papers, January. Swatman P.M.C. and Swatman P.A. (1989) Electronic Data Interchange - Implications for Industry, Proc Conf "SOST ’89", Terrigal, NSW, April.

Swatman P.M.C. and Swatman P.A. (1990) The Strategic Importance of Integrating Electronic Data Interchange into the Organisation, University of Western Australia Working Paper, July. Swatman P.A., Swatman P.M.C. and Everett J.E. (1990) Stages of Growth of an Innovative Software House: an Additional Criterion for Software Package Selection Australian Computer Journal, August. Tate P. (1989) Europe’s IS Execs Predict the Future, Datamation, March 1, 76:1-76:2 Tinsley T. and Power A.C. (1990) Why IS Should Matter to CEOs Datamation, Vol 36, No 17, September, 85-88. Walsh D. (1989) Spreading the Word Internally - How to Gain Staff Commitment, Proc IDC Conf "EDI - The Key to Profitability in the 1990’s", Sydney, NSW, December. Weber R. (1989) Controls in Electronic Transfer Systems: A Survey and Synthesis, Computers and Security, Vol 8, No 2, 123-127. Wilkinson R. (1989) Electronic Data Interchange and its Relevance to Australian Railway Organisations, Westrail Position Paper, June. Wilmot R.W. (1988) International Trends and Developments in EDI, Proc Conf EDI ’88, London, November, 15-21. Winston P.H. (1984) Artificial Intelligence, 2nd edition, Addison Wesley.