INET Conferences




Other Conferences

[INET'98] [ Up ][Prev][Next]

The Internet and EDI

Dick RAMAN <>


There is quite a lot being written and said these days about the Internet and electronic commerce. What's surprising about this is that no one ever mentions electronic data interchange (EDI) in these discussions, while just about all the large and medium-sized companies have been doing business with each other electronically for years via EDI. It seems as though the Internet world has no idea what EDI really is and what it can mean for electronic commerce.

In nearly all the material published about the information superhighway, the focus is exclusively on how consumers will sooner or later be able to obtain every possible product and service electronically without having to leave home. One never hears of the trouble that companies will have to go through to deliver them. Speculation runs wild over how much companies will have to invest in order to make it technically possible for consumers to take advantage of these services, but no one says anything about how companies will have to relate with one another in order to operate efficiently on the "highway."

It's high time that the EDI organizations in the world let their voices be heard and make clear that EDI is actually the backbone of electronic commerce.



There is quite a lot being written and said these days about the Internet and electronic commerce. What's surprising about this is that no one ever mentions Electronic Data Interchange (EDI) in these discussions, even though just about all the large and medium-sized companies have been doing business with each other electronically for years via EDI. It seems as though the Internet world has no idea what EDI is and what it can mean for electronic commerce. At the same time, largely due to all the excitement about the Internet, companies who have already started with EDI are becoming unsure whether the investments of the past years were really justified.

In nearly all the material published about the Information Superhighway, the focus is exclusively on how consumers will sooner or later be able to obtain every possible product and service electronically without having to leave home. One never hears of the trouble that companies will have to go through to deliver these products and services. Speculation runs wild over how much companies will have to invest in order to make it technically possible for consumers to take advantage of these services, but no one says anything about how companies will have to relate with one another in order to operate efficiently on the Superhighway. Even Microsoft's Bill Gates, in his book The Road Ahead, goes on for page after page about the science-fiction-like possibilities that will be available to consumers once the Superhighway is here. However, he describes EDI (which according to him stands for Electronic Document Interchange) in only 10 lines as something "unsuitable for ad hoc communications, although many companies are working to combine the benefits of EDI and e-mail into a single system." This declaration is typical of how the major players in the software market look at EDI.

The complete lack of understanding in the Internet world of the role that EDI can play in electronic commerce is a great danger for the companies which, at the end of the day, have to deliver the products and services. It is high time that the EDI organizations in the world let their voices be heard and make clear that EDI is actually the backbone of electronic commerce.

What is EDI?

EDI is the structured exchange of information between applications in different companies. It is about the relationships between companies. In a world that's getting smaller and smaller, and where even the smallest company can display its products before a worldwide audience, the interdependence of companies grows constantly. Small, isolated companies cannot survive in a market that spans the globe.

It is precisely through EDI that companies get the opportunity to coordinate their internal applications so the information stream flows smoothly between them. It has been clearly demonstrated over the past decade that this is not a simple matter, but we are now sufficiently advanced so that the benefits of EDI are clearly visible. Companies are beginning to realize that the structured exchange of data is the only way to plug into one another's business systems.

After all, all companies work together to deliver what consumers need. The vast majority of companies, however, have only limited contact with consumers; they derive the greater share of revenues from delivering products and services to other companies, which then deliver their products more efficiently to the consumer. It is logical that the attention today is aimed especially at consumers and satisfying their needs via the Superhighway -- that will determine how the whole production chain will have to work. There won't be very much change in the short term for companies that deliver to one another, though. As an example, it is clear that secure payment methods are not a problem among companies -- companies simply send an invoice and pay in thirty days. They will have new tools at their disposal to improve their production processes, but the main fury of the Information Superhighway storm will pass them by. We shall discuss the possibilities offered by the Superhighway shortly, but first, let's concentrate on EDI.

The slow implementation of EDI

The great concern of suppliers of EDI software suppliers, such as EDI-TIE, is "Why haven't the projections predicting an explosion of EDI been realized?" On the basis of the projections made by our clients five years ago, EDI-TIE should have revenues 10 times higher than is the case now -- even considering that EDI-TIE is the market leader in the Netherlands and has grown by five times in manpower and revenue over the same time. Apparently, it wasn't possible to implement EDI in companies at the speed which they themselves foresaw. It is difficult to determine what the reasons are, though a couple of elements are clear.

The big problem with EDI has always been that it is neither visible nor tangible. An EDI demonstration is essentially pointless. At best, you can show how the various functions are configured. The conversion process itself is invisible. Products that are in demand and quickly establish a market position, however, are not only high-tech, but also "high-touch." You can see them. You can touch them. EDI lacks that "high-touch" element which, for example, made the fax into a commodity in no time. In a desperate attempt to add visibility, many EDI software houses have resorted to adding bells and whistles rather than focusing on real functionality.

Another reason for the slow implementation of EDI is the lack of attention to the exchange of basic product information among companies. About 10 years ago, companies started with exchanging simple orders electronically. Subsequently, suppliers began sending invoices, which was the first thing they wanted to automate anyway. As time went by, companies realized that EDI actually had to be the "encoded" transmission of data if it was to be automatically processed by the receiver. And therein lies the problem: in order to exchange encoded data, the code has to mean the same thing in both applications, which requires the synchronization of the information between companies -- not a simple task.

It is surprising that companies didn't first concentrate on exchanging product family data before jumping into conducting transactions via EDI. It seems as though in the past, companies were so eager to get the benefits from EDI that they didn't take the trouble to first devote the necessary attention to exchanging the basic data. Companies are now starting to get bogged down in EDI exchanges due to that lack of product information in the internal applications. Only now is investment being made to make up for it.

Although the spread of EDI hasn't been nearly as wide as was predicted several years ago, the results achieved thus far have certainly been valuable, especially with the arrival of the Information Superhighway. In fact, the attention being paid to the Internet offers the possibility of showing where the power of EDI lies, while at the same time adding some sensible "high-touch" elements to it. It can also help solve a number of problems in EDI.

The essence of EDI

The essence of EDI is that data can be transferred from one application to another in a structured manner. From a technical point of view, this is quite easy to do. Any file in a previously agreed-upon format can do it. The matter becomes rather a bit more complicated when a company wants to send the same file to all its business partners. Then you have two problems: first, formatting the file, and second, coordinating the content to meet the information needs of all the partners.

The first problem cannot be solved by letting market forces have their head -- that would lead to every supplier needing a different format for every client. The customer is king, after all, and can impose his format on the weaker party. The solution is to have a generally accepted format, laid down in an International Standards Organization (ISO) standard. This format is actually a syntax in which different types of messages can be written. In Europe, the most common standard is UN/EDIFACT, and while there are some national standards, many countries are busy switching over to this United Nations standard. In the United States, ANSI X12 is the local standard, and still the most used, but the UN/EDIFACT looks as if it will clearly come out ahead as the standard for international traffic.

The second problem -- the content of the message -- is less easily resolved. In order to know which data elements a message must contain in the exchange of data among companies in every business sector in the world, it is necessary to make an extensive model of each transaction. This is where hundreds of specialists have been busy for years under the auspices of the United Nations. They describe in United Nations Standard Messages (UNSMs) what the content of a transaction must be in order to be universally applicable. The knowledge contained in these UNSMs is of crucial value for electronic commerce.

The weakness and the strength of EDI

From a software point of view, the EDIFACT syntax is just about the worst possible way to exchange data; computers are able to process much more compact and flexible formats. The EDIFACT syntax is actually based on the idea that one should be able to exchange messages by telex. That thought has clearly been overtaken by technology: today one wants to be able to exchange not only text elements, but also drawings, sound, photographs, and even video. This illustrates the weakness of the EDIFACT standardization process: people have been trying to figure out for a long time how binary elements can be included in an EDIFACT message, while much more powerful formats have already become the de facto standard. Microsoft, with developments such as Object Linking and Embedding (OLE), has already determined the direction in which we will be able to exchange information between applications.

The importance of the EDIFACT standard lies not in the syntax but in the knowledge embedded in the standardized messages, the UNSMs. EDIFACT should really be the "standards standard" in which the content, and not appearance, is established. Unfortunately, the field of vision for most software developers is limited to their own applications and they don't realize what demands their business partners will make on their applications when all companies are linked via networks.

Applications of the future will be individual modules which interact with one another, both inside and outside a company. Already, applications such as SAP, Baan, Microsoft Office, and Lotus Notes are a collection of individual modules brought together in a particular way on a user's PC to allow the user to do his or her job. A colleague can have the same package of modules set up in a completely different way because his or her function is different. It is quite conceivable that the data for the one module (and the module itself) will reside on a server within a company, while the data for the other module resides on a server somewhere on the other side of the world. This is the idea behind IBM's vision of Network Centric Computing: the network plays such a central role that the user no longer has any idea where the information is stored.

The trend is that data remains as close as possible to the source and that users access it whenever they need it. Only in this way do they have the guarantee that the information is up-to-date. The active dissemination of information has the drawback that the sender doesn't know exactly whether the receiver needs the information at that moment. The amount of junk mail we receive is the best evidence of this. When users have the guarantee that they can always find information on the Superhighway whenever they need it, the need to squirrel data away will disappear. This does mean that we need advanced search methods and intelligent "agents," helper applications that can independently search for bits of information.

Applying EDI on the information superhighway

Cable television companies will shortly be offering their subscribers so-called "set-top boxes," which will enable consumers to order products electronically, among other things. The cable company collects the orders and passes them on to a supplier, who takes care of the delivery to the customer. To this end, agreements are made over how the information will be electronically passed on. When a second supplier joins, the parties will again sit around a table to determine the contents of the files. With the third supplier, the same process occurs, and so on. In doing this, they are actually repeating the whole standardization process! In EDIFACT, UNSMs already contain the information the parties should be exchanging.

There is a clear role for EDI on the Superhighway. When discussing communication between applications, one's first thought should be "EDI." When a computer must be able to process the transaction, EDI is the only way to do it. When someone goes through an electronic catalog on the Internet to order products, he or she sees the information in HyperText Markup Language (HTML). The basic data such as packaging, number of units, and other technical specifications are given in text, and, if possible, the catalog is enriched with pictures, images, and sounds. This gives humans the possibility of processing the information, but the same information must be in EDI format for an application to process. When someone decides to order a product, a chain of events is set in motion.. In the first place, the product supplier must process the order. The supplier really needs to know only which article number is involved, the quantity, and who the client is. Secondly, the in-house application needs to know that the order has been made -- internally, several things have to be taken into account: warehouse space must be made available, Shipping & Receiving must know, and accounting has to be made aware. EDI is the only way to do this well.

This means that information on the Internet must be available not only in HTML but also in EDI format. If electronic catalogs are compiled on the basis of the data model incorporated in the UNSM, it is guaranteed that they are universally applicable. The EDI format itself is of secondary interest, as long as it contains the same information in a structured, generally accepted format. Until a better alternative presents itself, EDIFACT will serve here quite nicely. Internet servers should be equipped with the possibility of generating not only HTML, but also EDIFACT. A lot of products are now becoming available which can generate HTML. A flexible EDI converter such as the EDI-TIE Translate & Construct does this as well and of course generates the EDIFACT, too.

It seems obvious to send orders and other messages via Internet to the supplier, but there we run into a number of difficulties with the Internet. The problem on which the whole Internet world is now focused seems to be security. The Internet is, quite simply, not secure -- anyone can intercept and read all the data that can be sent via the Internet relatively easily. There is virtually nothing that can be done to improve the situation. It is inherent in the structure of the Internet. Those searching for a solution believe they have found it in the encryption of the file by the sender in such a way that only the recipient can decipher it. A number of systems have been developed to accomplish this, based on complicated mathematical algorithms designed to withstand years of computer analysis. Time after time, though, the exponential increase in computing power has meant that coding systems once thought secure have been cracked. This problem will one day, once and for all, be solved, but that is still ahead of us.

A second problem is the reliability of the connection itself -- there is no guarantee that a message sent via Internet actually arrives. The Internet is a collection of linked networks. There is no "owner" of the Net -- no single organization can take the responsibility of ensuring that a message arrives. For business applications, this is unacceptable; companies which exchange important business documents need the guarantee that messages arrive in a timely fashion. That is the strength of the Value Added Network Services (VANS) which have been used for years for EDI. A VANS provider can make a "Service-Level Agreement" with the users, in which he or she takes responsibility to deliver all messages securely, completely, and in a timely manner to addressees.

Market structure

Over the last decade, many companies have established electronic links between their applications. The Internet has become quite popular. This first of all drew much more attention to doing business electronically; however, nobody seems to know how. There are so many different ideas that nobody knows which one is the winner. In the press, EDI seems to be discarded as "Legacy" technology that did not make it in the past. And this is where the press is wrong: EDI is not a technology, but a business-driven way to make connections between applications in different companies. Whether the Internet or VANS are used to make the physical connection, the problem of aligning business processes remains. This is something that the bigger companies understand and are now more than ever going for. To them the Internet is rather scary: it proposes that they do business in a new way, using unproven technology that is unreliable and -- most of all -- not secure.

The electronic commerce market today is very much in its early development. A few Visionaries are picking up innovative products as breakthrough technologies that will bring great savings to the Visionaries' companies. EDI used to be in this phase of the Technology Adoption Life Cycle (see figure below), but it has moved into the mainstream market. EDI is no longer attractive to the Visionaries; the Internet has taken that role. No longer are benchmarks and product reviews important, and one doesn't see these appearing in the press anymore. Instead, EDI products just entered the mainstream market and are thus taken for granted. The core product is no longer important, but the services around them are. Now the Pragmatists in the market have taken an interest in EDI as something stable, with proven technology, international standards, a broad support, and a strong endorsement by the industry analysts.

Figure 1. The Technology Adoption Life Cycle (Source: Geoffrey Moore)

The early market

It will be the Visionaries who will pick up different types of products to accomplish their objective. The Visionaries are people who are driven by their own instincts and self-assurance; they take a greater interest in technology than in industry and are not very price-sensitive -- what they hope to accomplish with the product far outweighs its price. They look for products with unique functionality and state-of-the-art performance and architecture. They need highly driven "guru"-type salespersons to sell them this product. In fact, Visionaries are so sure of themselves, that once they believe in a concept, the product that might bring its implementation may be almost a prototype -- as long as it does the job.

Electronic commerce products fit this description. These products are still in a "prototype" stage, and they need strong technical innovative attention to turn them into the kind of product the Visionaries want. A lot of resources will have to be devoted to developing the state-of-the-art electronic commerce products. What these products need is to get a high profile and to be internationally accepted. What is needed are generic Internet-based products that have the potential to be used by many companies. This type of product will enable us to move electronic commerce into the mainstream market when it reaches that stage in the next 2-4 years. This is a particularly difficult transition, as is clearly described in Geoffrey Moore's book Crossing the Chasm.

The mainstream market

EDI is already in the mainstream market and it needs to be positioned with a market-driven approach. Here the potential buyers are Pragmatists looking for suppliers with a large installed base, supported by third parties and products that are the de facto standard, that have a high quality of support, and for which the cost of ownership is low. The core EDI product is no longer looked at separately; instead, buyers are looking for a "whole" product, where the generic product is augmented by a variety of services and ancillary products.

Today's mainstream high-tech market is different from what it used to be in the sense that buyers require a much more individual approach. The days of mass marketing and mass production are over -- buyers do not want to explain their requirements over and over again, but will go to the supplier who understands what they want.

Support revolutionSupport evolution
Break away from the packStay with the herd
Follow their own dictatesConsult their colleagues
Take risksManage risks
Motivated by future opportunitiesMotivated by present problems
Seek what is possiblePursue what is probable

The fact that EDI is in the mainstream now requires a market-driven approach. For electronic commerce this means it must be clear that this is the way business is done. This must be clear to the Pragmatists, who are basically willing to do whatever will save or bring them money, without having to take risks.


Cyber Assisted Business-EDI (CAB-EDI) is a concept for a multifaceted electronic commerce system with EDI in its rightful place as the backbone of the system and utilizing all the electronic communication tools and technology that will be available on the Information Superhighway.

The business user must be able to send messages via the Internet or a VANS or an X.400 system with equal ease. An access point, which allows the user to communicate with both systems over the same physical connection, would offer this capability. The Internet and the VANS would need to remain completely separate in order to guarantee security and reliability. The choice is then up to the user to determine whether he wants to send a particular message securely and reliably but at a cost, or whether he considers the security and reliability of the Internet adequate to send his message virtually cost-free. It is obvious that important messages (both EDI and e-mail) should be sent through the VANS and that the Internet should be used for less critical applications and for downloading large files.

Under the CAB-EDI concept, the Internet would be used to solve the age-old problem of keeping basic product data files up-to-date. The files are supplied by the producer of the product and can be accessed in a structured format that can be processed by the in-house application. There are two ways in which this information can be found. The first is that one goes looking for it and finds the information with the help of a powerful search engine which seeks out where the files can be found in HTML in order to later download the information in EDI format. The second method is to use an EDI-Agent that is configured to go at a predetermined time to a particular Internet address and there to check if the available information has been changed since the last visit. The EDI-Agent is intelligent enough to know what to do when the information has been processed: it can offer the information -- after having done the EDI translation -- to a given in-house application, alert a human operator, send a message, and much more.

It is especially the ability of the World Wide Web to make hypertext connections between files that offers untold possibilities for CAB-EDI. By maintaining their product information on their own system, suppliers can use this sort of hyperlinks to set up a Virtual Information Warehouse to lead clients to the right supplier and the right product. This is a nightmare vision for the many suppliers who are doing everything in their power to avoid the possibility of clients shopping to find the supplier with the lowest price. This may be a danger with consumer products, but many other factors -- guarantees, delivery terms, and service -- play a much bigger role in the relationships between companies, so there is less danger there. It is, however, of the greatest importance that this kind of information is available; it is a precondition for efficient transactions between companies.

Although EDI is specially designed for communication between companies, there is already a lot of experimentation with EDI in the relationship between consumers and suppliers. When consumers place orders, EDI offers the possibility of letting suppliers receive the orders in a structured way so that they can be immediately processed. The consumer has no idea he is using EDI and the supplier only has to implement one method to handle incoming orders. Powerful Query-Response applications offer the possibility of doing this and will take on great importance in the future.


With the explosion in the Internet, we are getting a vague idea of what the world will look like when the Information Superhighway is a reality. The role of EDI in this world is crucial. Electronic commerce will be significant for business-to-business only when EDI is an integral part of it. EDI is really the backbone of electronic commerce, though this appears to be not yet clear to the Internet world. EDI organizations have the mission to make this clear at all levels in our society. For companies, it means they must grab the opportunity now to become familiar with EDI, in order to be ready for the arrival of the Information Superhighway. The investments that have already been made in setting up EDI relationships will pay off in the future. The introduction of ideas such as CAB-EDI will enrich efforts already made in EDI with an untold number of attractive applications which will lead sooner to an enormous increase in the importance of EDI than to its demise.


  • Cyber Assisted Business by Dick Raman (ISBN 9090100024)
  • Crossing the Chasm by Geoffrey Moore (ISBN 0887307175)


Dr. Theodorus Hendrikus (Dick) Raman graduated from the Free University of Amsterdam where he studied Business Economics. After a brief teaching career, he started working for IBM in The Netherlands where he was trained as a consultant for the Information Network department. He was also assigned to the IBM-IN development department in Warwick (UK).

In 1987 he left IBM to start his own company, providing software for Electronic Trading. His company, EDI-TIE BV, has grown considerably and is the main European provider of standard EDI Software. Today EDI-TIE has five Subsidiaries and a growing number of Agents worldwide.

Recently he has gained worldwide recognition as an expert in the field of Electronic Commerce. He has published a successful book as well as a number of articles on the subject and is a recognized Visionary and Keynote speaker on EDI/EC events worldwide.

He has been the technical advisor for the Dutch government for many years in the UN body that governs the simplification of trade: the WP4. He now holds a position in the EC Working Group of CEFACT, the successor of the WP4; and as Board Member and Chairman of the EDI Committee of the EEMA, he is also the official representative of that organization in CEFACT.

Postbox 538
The Netherlands
Tel: +31-23-563-9971
Fax: +31-23-564-1748
Home Page:

[INET'98] [ Up ][Prev][Next]