It’s a technology that works and, in the risk-averse 21st Century, that’s what counts. EDI is, at least for now, alive and kicking.“It’s like the Energizer Bunny,” says Rico Giovannini, vice-president of sales for Mint Technologies Inc., a Toronto-based electronic payment integrator. “It just keeps going and going and, if anything, we find it’s going stronger.”
That’s probably a big surprise to the e-commerce pundits whose favourite pastime for the last decade has been predicting EDI’s imminent demise.
In 1997, Ethernet inventor Bob Metcalfe pronounced it “terminally clunky.” That same year, William Schrader, CEO of the commercial Internet provider PSINet, famously announced that EDI would not live past 1999. Extensible Markup Language, or XML, was supposed to revolutionize the supply chain and serve the coup de grace to EDI.
“When XML came out in 1998, it was supposed to blow EDI out of the water,” says IDC Canada analyst David Senf. “That didn’t happen.”
It might happen yet, but eight years later, EDI is still here. PSINet, it’s worth noting, is long gone. “I was one of those people who thought that EDI was dead,” says Robert Hall, CIO of the Speedy Transport, a Brampton-based national trucking and logistics company. “But it seems to have resurrected itself. We have a lot of customers who want to use it.”
It might be overstating the case a bit to call it a “resurrection,” though Hall says he is considering hiring an EDI specialist to deal with all of the requests to use the technology. On the other hand, Senf says, it never really went away.
“It’s here to stay, at least for now,” he says. “And if you’re creating EDI transactions already, and you’ve got the applications already in place, you’ll probably want to get your partners to use it.”
EDI has been a critical supply chain technology since before most companies knew that they even needed to have a supply chain.
It was developed in the 1970s and 1980s by the bigger companies in the automotive, transportation, consumer products, grocery and retail industries. The idea was that, with a standard for computer-to-computer transactions, companies could take human error and data-entry lag out of the transaction and scheduling process, and run a seamless, automated supply chain.
On the other hand, EDI is far from perfect. In its heyday, it was never quite seamless, and the automation required a major investment in then exorbitant computers and equipment.
The biggest problem all along was that it was extremely inflexible and resistant to change. Users were stuck using pre-defined transaction sets, rather than being able to adapt the transactions to their needs. “That has been a big handicap,” Senf says. “If you use EDI, you have to deal with the reality of it not really being extensible. It’s all hard-coded, while XML is, by definition, extensible.”
Nevertheless, EDI pretty much defined business-to-business commerce in the industries that pioneered the technology. Despite EDI’s problems, says Info-Tech Research analyst Carmi Levy, users have been loathe to jump ship.
“Despite its odiousness — EDI has historically been proprietary, expensive, and everyone had a different versions — there’s been nothing to replace it until XML came along, and that’s just starting now,” he says.
“Companies are fundamentally risk-averse, and if you suggest that they rip out an infrastructure that they’ve been using for 25 to 30 years, they’re not going to be receptive.”
At the end of the day, Levy says, EDI works, and that’s what matters to these users. “It gets the job done,” he says. “It got the job done yesterday, and it will get the job done tomorrow. That’s all that matters to them.”
In fact, many of the companies that want outfits like Speedy to do EDI are among the original pioneers of the technology in the notoriously risk-averse retail and consumer products industries.
“We do a lot of business with retailers,” Hall says. “They’re the ones pushing it as much as anyone, as well as third-party logistics companies. But it’s really the guys like Wal-Mart and Sears who want to use EDI.”
Big retailers like these have enough influence to force smaller companies, like Speedy, as well as their first-tier suppliers, to get into the EDI supply chain.
And because the first tier is doing EDI, the technology has filtered outward until the volumes of business are small enough to be handled by phone or fax.
The emergence of new logistics technologies like radio frequency identification (RFID) have only reinforced the trend.
Last year, Wal-Mart told its Top 100 suppliers that they would have to be RFID-compliant by the beginning of this year and, though most missed the target, the pressure to conform to a mandate to implement automated logistics tracking systems has revitalized EDI. After all, RFID only provides real-time logistics data. According to Giovannini, EDI has historically been the medium by which that kind of data is shared, and the big retailers have rediscovered its usefulness.
“Some of the newer initiatives like RFID and data synchronization have a lot to do with it,” Giovannini says. “They give a whole trading community visibility into the supply chain, but you have to have the data interchange to document it, and that’s the role for EDI.”
Levy isn’t so sure. In fact, he believes that one of the reasons for the persistence of EDI has been Wal-Mart’s, and other big retailers’, continuing trials and tribulations with RFID. The logistical tracking technology is more likely to harm EDI, and not help it.
“The delays in implementing RFID indicate the immaturity of XML, and this has kept EDI viable,” he says. “As goes RFID, so goes XML. The two technologies are inextricably linked, and as long as XML doesn’t take off, there will be a need for EDI.”
Nevertheless, the EDI of the 21st Century is not quite the same thing as the technology pioneered in the last quarter of the 20th. It still relies on the same transaction sets — standard transaction templates — defined by the Accredited Standards Committee’s X12 (ASC X12) and UN/EDIFACT standards. But the medium of interaction has changed radically.
“The big change for us is that we’re not using the VAN (value-added network).” Hall says. “We’ve started sending EDI over the Internet. We’re able to do that because we’ve started to do Secure EDI over the Internet.”
This is symptomatic of a trend that Levy sees, for canny developers to find ways to get the old dog to do new tricks. In the last couple of years, there have been EDI-to-Web screen-scraping applications for small businesses, and even efforts to translate its transaction sets into XML.
“There is probably room for creative vendors to find ways to bring XML and EDI together,” Levy says. “But I would say that the long-term prospects for the technology as a whole is not good.”
While that might be true, Hall expects the conservatism of his customers in the retail business and the sheer utility of the technology, for all of its warts and wrinkles, to keep EDI alive and kicking for some time.
“EDI will be around for a long time.” Hall says. “It has gathered new life, and the standards are very much in place and people adhere to them. It’s a technology that works, and that’s what people are looking for.”
Just how long remains to be seen, however. Levy declines to make a detailed prediction.
“There’s no question that EDI will die,” he maintains. “It’s just a matter of time. Will it enjoy another 25 to 30 years?
I doubt it.”

Share on LinkedIn Share with Google+