Technology News Technology News



Another Perspective


HOW TO DOWNGRADE YOUR BUSINESS PARTNER

According to IBM reseller Glasshouse Systems, Big Blue promised it wouldn't give a rival the 21 percent extra discount Glasshouse was getting in the pending deal with financial services firm SEI.  Glasshouse appreciated the incentive and poured extra effort into its sales effort at SEI Investments.  Surprise! SEI jilted Glasshouse and signed with rival reseller Mainline.  Glasshouse took a beating.

The upshot, Glasshouse pleaded, was that IBM, that double-crosser, ought to make it whole.  Stuff and nonsense, said Honorable Anita BrodyThe Glasshouse claim fails; IBM prevails.  The case is now a model for users who want to break free of incumbent resellers.

Only part of the lawsuit that tells the Glasshouse-IBM story is visible.  Much of the record has been sealed, basically because the case necessarily included some confidential commercial information, so the court kept a lid on testimony and other evidence.  But, it turns out, the visible portions of the record provide considerable guidance to any IBM customer that feels uncomfortably trapped in a relationship with an IBM reseller.  Where the case does not provide answers it certainly helps raise penetrating questions.  And it tells the story of one customer that downgraded its incumbent IBM dealer.

The case in point, formally known as Glasshouse Systems, Inc.  v.  International Business Machines Corporation (Civil Action 08-2831), was heard in Philadelphia's United States District Court for the Eastern District of Pennsylvania beginning in June 2008.  The ultimate findings were issued three years later, in July 2011.  During the proceedings, witnesses to the events that led to the dispute and experts on the workings of IBM and its reseller network told the tale of a deal that went sour for one IBM reseller and ended up in the hands of another.  The tale revealed aspects of IBM's sales practices that were previously known only to the privileged companies IBM calls Business Partners, or BPs, the firms authorized to sell IBM systems on Big Blue's behalf.

Key among the practices highlighted during the litigation is IBM's issuance of special discounts that, as a practical matter, bind a reseller and a customer to a specific deal.  Without these discounts, it might cost end users vastly more to acquire computers, resellers might not work as hard to match the customer's needs and IBM's catalog, and IBM might not be quite as successful in keeping customers loyal to its brand and to particular server families.  The flip side is that customers who are unhappy with their reseller can have a very tough time switching to another IBM dealer because that new candidate might not be able to get the kind of discount the incumbent obtained.

Yet, as the Glasshouse case shows, if a customer is willing to express discontent, then there will be action beneath the veil of confidentiality surrounding IBM and its network of resellers.  If it wishes, IBM can change the terms under which it sells hardware to a prospective dealer, thereby enabling that dealer to offer goods to the user at prices similar to those the user would pay the incumbent.

IBM of course doesn't like to disturb the connection of a reseller to a user, but neither does it want a customer becoming so dissatisfied that it stops upgrading installed IBM systems or, worse, turns away from Big Blue and its BPs.

When Glasshouse Systems first began talking to SEI nearly ten years ago, in 2002, SEI was running Amdahl mainframes with IBM systems software.  For the next four years Glasshouse worked to turn SEI toward IBM, initially by wining some peripherals deals and in March 2006 through migration to IBM mainframe hardware.  This was only the first step, however.  By the end of 2006, Glasshouse was trying to move SEI forward once more, this time to a newer generation of IBM mainframes.

The user didn't get much rest, and during 2007, as Glasshouse kept pitching an expensive processor upgrade, SEI became fidgety.  SEI secretly got in touch with another IBM reseller, Mainline Systems, a development Glasshouse only heard about after the process of defection was underway.

When Glasshouse first heard rumors about SEI and Mainline, it remained pretty confident, believing it held a key financial advantage.  Glasshouse had fought for special discounts on the hardware it was buying from IBM for SEI, including one big break that amounted to 21 percent of list.  This huge price cut was called ODP, for Opportunity Development Pricing.  IBM offers this break to a reseller that puts in an extraordinary sales effort.  .  .  and can prove it.

SEI seems to have been unaware of the ODP price break Glasshouse had obtained.  This is typical.  In general, customers have very little information about the prices that dealers pay IBM for equipment.  Customers only know what dealers ask for the machinery (and whatever else might be included in the transaction).

The way it works in practice, a dealer can pass along some or all of any price break to a customer or it can keep the difference.  In large deals, such as the one at SEI, a chunk of the money IBM lets the dealer keep will usually be used to support the dealer's help creating a migration strategy, planning documents, and so forth.  Basically, IBM is paying for the PowerPoints.  If the reseller pleases the customer and closes the deal, everyone wins: The user gets new equipment plus help with planning and migration, the dealer gets a lucrative sale, and IBM sells a bundle of hardware, software, and services without doing much more than booking the order.

Part of a dealer's calculation when examining an opportunity includes its assessment of just how likely it will be to win the deal that's under discussion.  Part of that calculation involves the dealer's sense of exclusivity.  If the dealer thinks it is the only vendor the user is seriously considering, that's one thing.  If the dealer feels it is in a race with other resellers pitching the same solution at about the same price, that's another.  In the SEI situation, that's where things became murky for Glasshouse.

Glasshouse thought, mistakenly as it turned out, that it had a tremendous financial advantage over other mainframe dealers and particularly over Mainline.  It thought so because Glasshouse believed Mainline was not going to get the 21 percent ODP discount.  Glasshouse thought the break it had obtained was unique in the SEI deal and that Mainline would have to pay IBM a lot more for the IBM hardware.  In court, however, it transpired that IBM never actually promised the ODP exclusively to Glasshouse, as the reseller believed.  And in the end Mainline, paying the same discounted price for hardware as Glasshouse was quoted, won the big upgrade deal.  SEI signed with Mainline toward the end of 2007.

Glasshouse not only lost the SEI deal, it lost its lawsuit, too.  This loss occurred even though IBM's behavior was imperfect to say the least.  Judge Brody found that IBM was vague about the way it handed out ODP discounts to rival BPs during the years on which the case is based.  (The judge said IBM subsequently made its practices more consistent and transparent.)

Now the reason this is so important is that Glasshouse went to court with a half dozen reasons it believed it had been illegally mistreated by IBM, but after running into IBM's legal buzz saw, only one line of argument survived.  The sole claim that was strong enough to merit a full trial, the court found, was what lawyers call promissory estoppel.

Here's more or less how it works.  There are two parties to the basic situation.  The first promises something to the second.  Relying on this promise, the second party acts to obtain the benefit of the promise and in doing so expends resources.  Then the first party reneges on the promise and as a result the second party gets hurt.  Under the doctrine of promissory estoppel, a court can compel the first party to stick to its original promise and to compensate the second party for harm caused by its disavowal of the original pledge. 

That's kind of abstract, so here is an example: A landlord promises to hold a tenant's rent down for at least the next decade if the tenant renovates the subject premises.  The tenant invests in the rented property but after a couple years, when the lease is up for renewal, the landlord boosts the rent and threatens eviction if the tenant doesn't pay the higher rent.  If the tenant can prove to a court that the landlord promised a rent freeze and if it can show that this promise is the reason the tenant spent money on the property, a court can order the landlord to revert to the original deal (or compensate the tenant in some other fashion).

Boiled down to its essentials, the Glasshouse case was turned into a textbook dispute hinging on promissory estoppel.  And unfortunately for Glasshouse, IBM's attorney mounted a defense quite worthy of classroom presentation.  The lead lawyer for Big Blue was Robert Feltoon of the Philadelphia firm Conrad O'Brien.  Feltoon is an alumnus of New York's Cravath, Swaine & Moore, IBM's principal outside counsel and arguably the best law firm in the country.  (Cravath considered Feltoon top notch and on another occasion trusted him to fight the first stage of a complex dispute involving a claim of conflict of interest on Cravath's part; Feltoon succeeded.) Glasshouse had turned to another major league Philadelphia lawyer, Tim Russell of Spector, Gaden, and Rosen, but as the case unfolded there was very little Russell could do.  Well before the three-year battle ended, it became clear that Feltoon would crush his opponent.

A pivot came after a Glasshouse executive swore IBM had promised it protection from rivals.  The protection came in the form of the exclusivity of Glasshouse's ODP discount, which, the witness avowed, would protect Glasshouse for at least 90 days.  But IBM personnel testified to the contrary.  The court did not find that Glasshouse had lied but rather that it had been mistaken.

The second step in a classic PE strategy involves proving that the claimant acted to its disadvantage because it believed the foundation promise.  Glasshouse tried its best to show how its sales effort was scaled up because Glasshouse thought it had an exclusive lock on the ODP.  IBM argued that there were no facts supporting this contention.  The testimony governing this aspect of the case has been sealed.  The details are hidden, but somewhere along the line Glasshouse apparently failed to persuade the court that its version of the story held water.  (This is one aspect of the case that might have been challenged on appeal, if, Glasshouse, like some of Judge Brody's critics, thought she was in error.)

Finally, Glasshouse had to show that its effort, frustrated when its reliance on IBM's purported promise led nowhere, resulted in harm.  Glasshouse could not establish a chain of causality connecting IBM's reneged promise and the costs of its own consequent activities.  Basically, Judge Brody didn't find that the claims made by Glasshouse were supported by the facts of the case.

Even if Glasshouse had won against IBM and collected the $3 million it asserted it was entitled to, it would not have been able to retrieve its relationship with SEI.  The user was so unhappy with Glasshouse that it actually paid $100,000 more for deal from Mainline than it would have paid to Glasshouse.

So, what emerges is that a dissatisfied user can unhook an IBM reseller and seek an alternative vendor.  IBM might be secretive about its arrangements with both the incumbent and alternative reseller, but the manufacturer is willing to support competition among its dealers if the user insists.

It doesn't matter whether the systems involved are mainframes, as was the case in this instance, or Power Systems boxes.  If a customer is really unhappy with a reseller and either goes out looking for a new BP or manages to get the ear of an IBM executive, IBM will help the customer fight for a comparable deal from a more acceptable supplier.

— Hesh Wiener August 2011


Copyright © 1990-2017 Technology News of America Co Inc.  All rights reserved.