If you are interested in B2B e-commerce, you are interested in the history and future of the EDI giants:  GXS and Sterling Commerce.  They remain behemoths in the space, despite almost 20 years of Internet-based encroachment.

Pre-Internet Days

EDI was the original language of B2B e-commerce.  It began as proprietary technology used exclusively by the largest corporations in the 1980s.   Gradually, EDI became more  standardized in the early/mid ’80s, and by the late 80’s and through the 90’s it was a very fast-growing business.  EDI business documents were transported over still-proprietary Value Added Networks (VANs) until the advent of commercial use of the Internet in the  ’90s.

Evolution of the Big Boys

Throughout this time and over the past 10 years, two companies evolved to be the leading EDI networks from among numerous competitors:  GXS and Sterling Commerce.  (The pattern is similar to the consolidation in payment networks and cellular networks.  Eventually network economics mean only a few networks can survive and compete.)

GXS was initially the EDI business of GE which was spun off and then bought several other network including IBM’s EDI business and, more recently, Inovis.  GXS’s business appears to have declined slowly and steadily throughout the 2000s, but recovered in 2010 and 2011.  GXS estimates full year 2011 revenues will be $478 million and will drive an eye-popping $155 million in adjusted EBITDA!   Clearly the business is still pretty impressive in throwing off cash!

Sterling Commerce grew from a $200 million business in 1995 to over $600 million in 1999 and was bought shortly thereafter for the princely sum of $3.9 billion by SBC.  (It too was very profitable.)  SBC then merged with AT&T and Sterling remained an AT&T subsidiary until last year when IBM paid $1.4 billion in cash for it.  Sterling appears to have remained at around $600 million in revenue, shrinking some, but then acquiring some other companies along the way (more on that later).  (By now the reader may have noticed that IBM sold its EDI business to GXS and then a few years later bought its larger competitor–we will come back to that as well!)

When the Internet took off in the late 90’s, many pundits predicted that XML and the Internet would kill the EDI behemoths in short order due to EDI’s relative high cost, inflexibility, and use of proprietary VANs.  As you can see, rumors of the death of the EDI behemoths were greatly exaggerated.  GXS and Sterling staggered a bit, but they survived and remained profitable businesses, even if they did not exactly thrive over the first decade of the new millennium.  They still represent the giants of B2B e-commerce with over $1 billion in revenue between them.

The EDI giants survived for several reasons:

• They were very profitable, so they could, and did, lower prices aggressively.
• Pundits forgot that “sunk costs are sunk”, but clients did not.  While EDI is expensive, it is mainly so in the set-up costs and less so in the on-going operational costs, so customers (appropriately) looked only at the marginal costs of using EDI when considering whether or not to replace it.
• XML and AS2 were new technologies that took a while to be trusted and take root.
• GXS and Sterling moved their attention to managed services (outsourcing) of their clients’ B2B e-commerce strategies, which was really smart because retailers and others do not want to maintain a core competency in a technology that is not even taught in schools any more!
• GXS and Sterling wisely convinced customers to continue to use the EDI message types and simply move some of the volume to AS2/ web-based transport.
  They became experts in facilitating this transition.

Each company survived, but neither really thrived, during a period when at least another $2 billion in B2B commerce revenue was generated by the likes of Ariba, GHX, Emdeon, e2open, and many others.  Why?  Most likely because of the Innovator’s Dilemma.  GXS and Sterling simply did not want to see their base business rapidly disappear through cannibalization caused by the new XML/Internet-based protocols.

Both GXS and Sterling made small acquisitions in the 2000s that suggested a desire to add value to the messages they were transporting.  (GXS bought Celarix, HAHT commerce, and even Rollstream very recently.  Sterling bought Yantra, Nistevo, and Comergent.)  But to a large extent, the EDI behemoths continued to see themselves as a Post Office –transporting the messages electronically–but not “opening the envelopes”.  They did not really try to understand what was in the messages they transported so they could add value or analytics to the data.

GXS and Sterling also nibbled around the edges of adding applications to generate or consume these messages on the buyer or supplier side of the network.  (To be fair, Comergent was a nice step on the supplier side for Sterling.)  And they made few attempts to verticalize, add financial services, or otherwise capitalize on their positions.

GXS and Sterling’s owners, (PE firms and AT&T respectively for much of this time) were careful to keep the cash flowing and limit the investment–which is a reasonable approach.  But to look at what the alternative might have been, it’s interesting to take a look at a company that started as an EDI network and evolved more rapidly and aggressively to add applications, value-added services, and analytics.  That company is Emdeon.

In the 1990s, the predecessor to what is now Emdeon was a $100 million dollar processor of largely EDI-based healthcare transactions for payers (insurance companies) and pharmacies.  Over the next twenty years, Emdeon kept its core business of connecting payers and pharmacies for transactions and used that cash, IPO cash, and investor cash to grow organically and through acquisitions.  Emdeon added:

• workflow/ revenue cycle management applications on top of the network for providers and pharmacies
• Analytics and reporting to provide real-time information to all players in the process
• And managed services and consulting services

Emdeon went far beyond being a post office for their customers and managing just connectivity to becoming integral to their workflow, their revenue management, and their operations.  The result by 2010 was a $1 billion company with connectivity representing only 20% of revenue (from 100%), applications 60%, services 15%, and analytics 5%.  (Over time Emdeon expects each area to be 30% of their revenue, except basic connectivity which will then be only 10% of revenue.)  Emdeon was recently taken private by Blackstone at a valuation of $3 billion or about 3x revenues, a nice premium to the Sterling transaction.

Summary

It would have been interesting to see GXS or Sterling take the Emdeon approach in the late 1990s and 2000s, rather than the more conservative tack they took.  One or both of them might be truly huge now.  It seems that GXS will continue to take a conservative approach, as it carries a lot of debt from its private equity owners.  But who knows what IBM will do with Sterling?  IBM has a lot of software assets that can be leveraged and certainly knows analytics and services.   Perhaps it is not too late to see one of the giants awake from their slumber!?  Who knows, maybe this is why Warren Buffett invested in IBM? 😉