Interconnection as an Anticompetitive Weapon

Consolidation of the Value Added Network Market

in re Loren Data Corp v. GXS Inc.

Alan D. Wilensky, Analyst,

bizQuirk Market Strategy

Author’s Note: within seconds of posting this article, I received email asking me for links to Loren Data Corp’s web properties, so here: (API docs)

An innovative B2B network communications company is struggling to survive its decade-long dispute against the largest company in their sector. We might cast Loren Data Corp. as the Rebel Alliance, and GXS, Inc. as the Empire, but that would make me nerdy. This dispute properly concerns interconnection policies: how do competing networks cooperate by way of interconnections, in order to create functional markets that attract, serve, and retain their subscribers?

We instinctively know that our calls, emails, and each web site we visit makes its way to our homes and mobile devices via an amazing mesh of independent, competitive networks.  We never know what network our customers or family members subscribe to, nor should we – we just dial, search the web, or hit a send button…’s magic, it really is!

Networks must compete with each other, yet also collaborate via interconnections in order to indulge our illusion of seamless communicating; the fact that we don’t have to think about it speaks volumes about how far  these connected industries have come, and how far networking technology has evolved.

There is no doubt Interconnections are important. But, what happens in unregulated markets when things go awry, when Private Equity rolls up the top three EDI Network Operators (VANs) in the sector, and then refuses interconnection  to “select competitors”?

Interconnections between networks are a very powerful competitive force. This is why certain markets deemed ‘too important to leave to competition alone’, were regulated under “Laws of Common Carriage”, roughly defined as:  1) carriers must accept all who apply for their services, and 2) the carrier must agree to interconnect with licensed, or otherwise bona-fide market serving peers.

The meat of the issue lies in understanding the antitrust arguments brought by Loren Data Corp, boiling down to GX’S denying an interconnection to its TGMS (Trading Grid Messaging Systems) VAN (Value Added Network). Loren Data Corp’s EDI Network operates under the Trademarked Brand, “ECGrid®”.

The question of recourse:

When a large network simply refuses to interconnect with a smaller competitor, what recourse does a smaller competitor have?

The answer depends on whether a market sector is regulated (i.e., telecoms) or deregulated (i.e., the Internet).  In markets that are subject to the laws of common carriage, the FCC provides avenues for petitioning and enforcement, as well as maintaining a system of “bureaus”, for monitoring and applying corrective actions when necessary. The Wireline Competition Bureau is such an office in the FCC that monitors, you guessed it, wireline telecoms.

EDI communications carried by VANs are presently not regulated by the FCC, although the most superficial research shows precious little consideration or due diligence was undertaken by the FCC when VAN communications were swept into the “unregulated” pile.

The telecommunications reform act of 1996 created a deregulated class of services – such as pure data communications, and most Internet services, (the Web, email, most VOIP, and peer-to-peer communications, such as Skype and IM, generally). The 1996 Reform Act sought to foster growth by lightening the regulatory burden that plagued the telecoms sector, and for the most part, it worked. But not for services predating the Internet, such as EDI VANs.

Providing relief via the regulatory channel would require EDI VANs to be placed under the rubric of the Laws of Common Carriage. The FCC would need to research the sector, and define the market parameters  (size, capitalization, current interconnection practices, past history of litigation), that would trigger the regulation of certain VANs. The entire process  would be kicked off by a petitioner, often an aggrieved party, or by an industry coalition of allied competitors. It would not be a speedy process, but in some cases, the agency can act more quickly than the courts. There are also hearings and appeals procedures for any opposing parties.

Creating new regulations for EDI VANs can be properly called a long-shot; the momentum at FCC  remains firmly tilted in the direction of further deregulation.

Deregulation falters when considering Hybrid Communications Services Predating the Internet

The race for the Agencies to encompass telecommunications carriers that offered data processing services resulted in a question of regulating “hybrid communications services” – defined as “services wherein the data processing capability is incidental to the message-switching function or purpose.”

Nothing better describes EDI Communications over VANs. EDI messages travel from party to party, carrying valuable, time-sensitive content. The electronic business documents in their EDI format, such as invoices and purchase orders, are exact analogs of their paper counterparts, yet perform a more demanding service within modern supply chains. The drive to economize manufacturing and retail inventories, known as “just in time”, and “lean processes”, are wholly dependent on the electronic, real-time nature of EDI supply chain transactions.

The recourse arising from recognizing EDI VANs as hybrid communications services is bolstered by the concept of Bailment. Bailment occurs when a third-party is paid to convey an item(s) of intrinsic value to another party. Bailment is fundamental to the laws of common carriage.

EDI messages are mission-critical, real-time services that businesses depend on.  Retail supply chains and manufacturing procurement systems depend on EDI for system to system order-management and trading-partner community integration. We are talking about trillions of dollars of trade purchasing and supply side fulfillment , and VANs are the medium that represents the movement of this trade.

In the context of Loren Data Corp v. GXS, the regulators  simply need to realize that EDI is important, and that most EDI communications are business documents (invoices, purchase orders, and shipping advisories)  traveling over VANs, and VANs take bailment of EDI messages from the sender, and convey the unaltered message to a recipient. Postal Mail, overnight couriers, and freight companies are all governed and licensed as common carriers subject to bailment; EDI is the equivalent of Telex orders sent between computers. EDI is not an enhanced service.

The regulatory formula for declaring PE funded VANs that are “offspring of Consolidations”as the first candidates for common carrier regulation.

A regulatory formula for petitioning might be constructed as follows: EDI networks (VANs) are Point to Point message switching services, taking bailment of the valuable, time sensitive messages from sender to recipient. EDI is essential to the conduct of mission critical trade communications, with capital orders for goods. Mistimed, delayed, or repudiated electronic orders represent the risks associated with EDI in modern supply chains.
EDI messaging is therefore a hybrid communications service subject to laws of common carriage and bailment.

The largest VAN, GXS, funded with Private Equity, consolidated the subscriber populations and aggregated interconnections of the previous top three VANs.
GXS should be the first Value Added Network considered for regulation as a Common Carrier, and should furthermore be mandated to adhere to interconnection practices designed to foster fair competition within the sector.

The supply chain integration market, served by the VANs, is best served by a wide spectrum of competitive EDI providers who actively innovate and vie for all segments of the market – courting the large enterprise all the way to small businesses that are eager to EDI enable.

Failing Regulatory Relief, Recourse falls to the Sherman Antitrust Act

This is mainly a discussion of network interconnection policy. When regulation fails, the intersection of competition and law bring us to the Sherman Antitrust Act. There is a growing judicial record on Network interconnection practices and competition: these are so-called ‘network effects’, whose effect on competition is real, and almost always involves interconnection disputes between competing networks.

Network Effects (from Wikipedia) “When network effect is present, the value of a product or service is dependent on the number of others using it – also, the number and quality of the resources attached to a network (interconnections to other networks, a related influence of network effects)”, example:

GXS Inc purchased the top three VANs, therefore the trading partners attached thereby are amplified in value due to network effects. Alternatively, anyone using EDI or providing EDI services MUST cross the GXS border.

The Central Dispute

The plain reading of the dispute is GXS’ refusal to provide reciprocal message routing between TGMS (Trading Grid Messaging System), its preferred VAN property, and Loren Data Corp’s VAN, ECGrid®.

VAN interconnects are ‘software configured’ routing policies for delivering message traffic between competing VANs. The VANs provide electronic mailboxes and supporting services that facilitate commercial trade. VANs are best conceptualized as private email systems used for sending standardized messages concerned with the buying, selling, and shipping of goods between trading partners, and much more.

This dispute reeks of anticompetitive musk – a smell wafted by a panicked C-Suite that is plum out of fresh ideas. In my opinion, as an analyst specializing in the B2B tech trade, there will never be a GXS IPO.

In my personal opinion , GXS fears Loren Data’s founder, Todd Gould, whose knowledge of EDI Network architecture is expansive, and whose ability to innovate is limited only by the speed he can code and debug. Todd is just hitting his stride, and has innovations aplenty in his pipeline. GXS could only wish to approach Todd Gould’s fecundity in their EDI Communications products.

The Size Differential:

GXS is backed by Francisco Partners, a private equity fund in San Francisco (others funds have minority holdings). With 2800 employees, GXS, Inc.’s gross revenues hover around $450-500M/year.  GXS files forms 10K and 10Q with the SEC. For more details, search the Edgar System @

Loren Data Corp. is, by any measure, a small business. With a twenty-five year tenure in the EDI business, the company has gross revenues that waver about the $1M/year mark. Loren Data has five long-term employees and contracts with this author as an industry relations analyst.

The company is a lean, bootstrapped entity with an enviable low-cost of operations. Loren Data Corp could have certainly grown its revenues by a conservative 3-5X, had it not been thwarted by GXS routing issues, as well as being hampered in its line of business by GXS actively damaging its reputation, starting as far back as 2002.

Note: In spite of all of the damage attempted by GXS, in all of its forms, Loren Data Corp ECGrid has never dropped one Byte or character of message data – never. The company maintains functional routing arrangements that  handle data bound for any Trading Partners on GXS Networks . However, “functional” is not “equal”, especially in a competitive sector, and certainly not in the way that GXS offers standard VAN Interconnect reciprocal message routing to networks  with fewer subscribers and less traffic than ECGrid.

The lack of a native TGMS interconnect is costly, damaging to the company’s otherwise stellar reputation, and incurs additional labor.  Todd, an amazing network engineer, the very best of the best, sees to it that any routing deficits are always minimized via some very clever  (and totally unnecessary) engineering. The most damaging aspect of GXS actions are the ‘whispering campaign’. engaged in by their large support organization – whose megaphone amplifies the effects of routing problems, by pointing fingers at Loren Data Corp, as “not a real VAN”, as issues being the fault of ECGrid Netops staff, when the reverse is virtually always true. If anything, those in the know, know that ECGrid is the Cadillac of EDI Networks, and TGMS is a horrible, unreliable kludge.

For a deeper understanding of the precise damages wrought by GXS upon Loren Data via Interconnection Arbitraging of  TGMS routing policy , see the court transcripts or contact the author. There is so much more than mere data blockage at stake, and these actions by GXS are damaging to then entire EDI market, from end-users, to service providers, and consumers.

The Upcoming Appeal

Loren Data Corp’s initial Federal district court filing was not the company’s best moment. Although represented by a sincere and competent generalist attorney, antitrust cases within the networked market domain are best represented by specialists. The company’s first efforts were soundly thwarted by the District Court Judge, who stung the company by dismissing the majority of the Sherman Act Charges.

The lower court dismissal was brought to a well-known trial lawyer experienced in high-technology markets. Atty. Glenn Manishin confirmed the company’s lay reading of the Judge’s opinion – an extremely harsh rendering of the Sherman Act, favoring the monopolist, and thwarting out of hand the ability for a disadvantaged smaller party, Loren Data Corp, any opportunity to conduct discovery and prove their case.

The appeal reaches the 4th Circuit in Richmond, VA in September, if no other delays intervene. If Loren Data Corp and Atty. Manishin prevail, the case will be remanded back to the MD district court, and discovery will commence. But far more important, this will be a vindication of the basic charges originally presented against GXS.


The Fiasco began with GE’s selling its GEIS (GE Information Services) division to Francisco Partners as lead PE investor, and a pre-planned takeover of two major VANs: IBM-IE, and in 2010, Inovis. Globally, the Francisco PE juggernaut went on to buy ninety percent of the UK EDI market (as measured by message traffic), while Great Britain’s anti-monopolization regulator ignored the numerous and bitter complaints of England’s competing EDI Service Providers.

Loren Data Corp.’s founder, Todd Gould, was bidding to provide services to IBM’s IE VAN, even as its was in the process of selling out to GXS….even as the deal was being rammed through the regulator’s maw in 2005. Strangely, IBM must have had a bout of ‘buyer’s remorse’, and purchased the New 2nd place VAN, Mighty Sterling Commerce, now part of IBM WebSphere.

As a scrappy, typically entrepreneurial business existing solely by its founder’s foresight and creativity, Loren Data Corp. was just trying to compete against a newly formed GXS soaking in capital. Todd was asked, and provided, DOJ’s antitrust division with insight on the all but assured toxic results of the Inovis acquisition.

All was for naught, as the predicted deleterious effects have materialized, in spades—and now we are all so blessed as to have a bona fide, AT&T-style, toxic monopoly in the EDI communications business.

For journalists, the key point in covering this story reads thus:

The monopolization of the EDI communications market is not occurring because a PE fund (Francisco Partners) rolled up three competing value-added networks into GXS, Inc.  Rather,  it was the act of breaking a settled industry practice:  non-settlement routing among cooperating , competitor VANs.

The implications are manifold from a competitive, regulatory, and economic perspective. Primarily, without cooperative routing, the global network of VANs never would have materialized; interconnections made the market, and interconnections maintain the market. The same justification exists for the Internet – there is simply no  way to conceive of today’s Internet without peered interconnection.

Therefore, GXS breaking the cooperative interconnect model is more than mere competitive ruthlessness; its absolute proof that the leadership of GXS harbors nothing but contempt for the industry they deign to serve.

We simply take for granted all of our communications occur at any point in time over a mesh of competitive networks: Internet, phone, VOIP, e-mail, and even utility meters. Gaining control of networked resources, combining networks and buying up the subscriber populations – all an opening gambit in the game called ‘interconnection arbitrage.’ In this case of (GXS) refusing to honor data traffic of a smaller company (Loren Data Corp.), we see a prime example of network effects.

AT&T, once all-powerful, and with absolutely zero competition in long-distance services, purchased 100% of its nascent regional competitors. Any competitor would have to interconnect with AT&T merely to exist. AT&T had harnessed the power of the network effect before the term was even coined. AT&T filled its coffers with outrageous metered long-distance and local calling fees, all based on confounding time-of-day calling plans. Likewise, any EDI communications network simply has to cross a GXS routing border to exist – and therein lies the rub, and the seeds of potential FCC enforcement, and antitrust charges that survive the upcoming 4th Circuit Appeal.

It took a newly created FCC and the courts to battle to the brink of a verdict, and ending in a consent decree. Almost simultaneously, MCI won its antitrust appeal in the Supreme Court. A span of thirty years had elapsed in order to begin the divestiture of AT&T into the BOCs, the Bell Operating Companies.

While the key players in Loren Data Corp. v. GXS, Inc. are not quite as infamous as AT&T and MCI, lest one think that EDI communications are somehow trivial, consider that trillions of dollars in orders, invoices, and shipping advisories are communicated over hundreds of cooperative EDI VANs.

The FCC partially awakens from its Institutionalized Torpor, and the DOJ is certainly watching

After several detailed briefings to the proper regulatory bureaus at FCC,  as well as to the office of the Commissioners and the General Counsel, FCC may finally be warming to the concept that EDI communications over VANs are not simply enhanced Internet services, but are within the agency’s jurisdiction under the laws of common carriage.

The Remedy

GXS, Inc. needs to be divested into its original component companies, and forced to compete on a level playing field. The least relief that courts and regulators should grant is a consent decree, minimally forcing divestiture of the GXS VAN properties (TGMS particularly) into a standalone company operating under the laws of common carriage. Under such a scenario, GXS would become a) one B2B IT integration company, doing what it loves, and does best, i.e., managed services, and b) TGMS would be a separate entity, with the FCC baby sitting its ill-behaved former executives, and making sure that inter-VAN connections are treated with the proper, industry affirming respect they deserve.

In closing, I leave my colleagues with some open-ended questions for the comments section. If you are one of the rare supporters or paid shills for GXS Inc, and you would like to rebut my many diatribes, please do so below.

  1. Why is GXS, a giant EDI network operator, abrogating industry-accepted interconnection policies?
  2. Why would GXS use interconnections as a means of harming a competitor?
  3. Why do YOU think that there is no outrage in the market? From other market participants (competitive VANs)? The trading partners who are GXS clients? These companies are the unwitting heirs to a damaged market, when the sole VAN that caters to service providers is eliminated.
  4. Is the inverse possibly true? Is GXS averse to easy and cost-effective EDI communications services?
  5. How does Loren Data Corp. threaten GXS? We have reiterated ad nausea the differential in size, revenues, etc.
  6. Why did the DOJ and FTC, and even the U.K.’s Office of Fair Trade, close their collective eyes at this egregious, yet fixable act of interconnection arbitrage?