The Moral of This Story: Stay Out of CourtIn March 2000, a $10 million software company called Triple Point Technology signed three contracts with Transammonia, then a $1.5 billion petrochemical company, promising to link all 27 Transammonia offices with a state-of-the-art commodities trading platform called Tempest 2000. Triple Point also agreed to design and develop six interfaces between Tempest and Transammonia\u2019s existing PeopleSoft accounting system.For Transammonia, it was the quintessential Internet-era "business transformation project" that would launch the old economy company into the 21st century of Internet-based, real-time, 24\/7, free-flowing information. And the six interfaces between Tempest and Transammonia\u2019s existing accounting software were the crux of the deal. No point in having a free flow of information if it didn\u2019t freely flow into the general ledger.Triple Point assured Transammonia the interfaces wouldn\u2019t be a problem. In its business proposal to Transammonia from September 1999, Triple Point said no fewer than three times that it had "experience" building "fully integrated solutions." Triple Point could "custom-build any interfaces required," the proposal also said. The sides talked about the interfaces shortly after the proposal, and a month later, too, when Triple Point again assured Transammonia that it was "very familiar with the nature of these interfaces, having built them for some of our existing clients." Triple Point said it had built a similar Tempest-to-PeopleSoft bridge for Mieco, another chemical company.Transammonia was convinced. The company supplied a quote for Triple Point\u2019s press release?something about working efficiently in fast-changing markets. It was a happy nuptial, but then, nuptials are never unhappy. Marriages sometimes are.Failure Hits. Send in the Lawyers.But Triple Point, it turned out, was rather liberal in representing its experience. In fact, it had never developed the interfaces. The company had only managed subcontractors that developed the interfaces, and it had done that just twice.No one at Triple Point had mentioned subcontractors to anyone at Transammonia prior to the contract signing in March, but that was the plan. Triple Point hired PeopleSoft to develop the interfaces.But by August, Triple Point still hadn\u2019t provided PeopleSoft with a project plan and failed to monitor progress on the interfaces, other than through informal conversations. By the Dec. 31, 2000, deadline, zero of six interfaces was completed. In February 2001, Triple Point stopped paying PeopleSoft. A few days after that, Transammonia stopped paying Triple Point. Finally, Triple Point turned the screws on its subcontractor, setting a firm March 9, 2001, deadline for the interfaces to be completed.That deadline wasn\u2019t met. Nor were March 29th, April 24th or April 30th deadlines. Transammonia officially suspended the project on May 2, 2001. Within days, Triple Point executives were pleading with Transammonia to hang in there. Triple Point wrote letters to Transammonia claiming it was the subcontractor\u2019s fault for not devoting the talent or resources necessary to the project?despite the fact that Triple Point was so tardy providing a development plan and wasn\u2019t checking up on its subcontractor.The contract stated that Transammonia would get the six interfaces in 230 days at a cost of $375,000. But after 400 days, having been billed $635,000, Transammonia had one interface.So on July 11, 2001, Transammonia set a 30-day ultimatum: Make everything work or else. Triple Point could not do that. Transammonia refused to pay Triple Point any money it owed, so Triple Point sued its client for breach of contract, demanding $795,000 for services rendered. Transammonia then countersued for breach of the agreements and demanded its money back and then some.Thus divorce proceedings began.They didn\u2019t end until Sept. 25, 2003, more than two years after lawyers commenced billable hours and a full four years after Triple Point submitted its sunny business proposal.The New York Supreme Court\u2019s Judge Herman Cahn, in a 4,408-word decision from which the above narrative was taken, recounted the significant?that is, the ugliest?details of the affair. He ruled in favor of Transammonia. (Both companies declined to comment for this story, including Transammonia\u2019s CIO, Benjamin Tan. PeopleSoft also declined to comment.) But as victories go, this one lacked marrow. Transammonia will get a refund (a court-appointed special referee is still deciding the amount), but four of Transammonia\u2019s five counterclaims were thrown out. The one claim granted called for rescission?legalese for putting things back the way they were before the parties knew each other.Or, as the final footnote in Cahn\u2019s decision noted, "Transammonia asserts that it has paid over $1.5 million to Triple Point under the Agreements, \u2019and has nothing of value to show for it.\u2019"Save Millions with a ParagraphTechnology litigation?broadly defined as suing or threatening to sue over software project failures?is not new. The legal books are full of it. In one notable 1979 case, National Cash Register (NCR) was successfully sued by an electronics company called Chatlos Systems for breach of contract. Like Triple Point, NCR promised systems it didn\u2019t deliver. In 1991, service staffers for Wang Laboratories (now Getronics Wang, a branch of Dutch company Getronics), attempted to fix a deployment of computers they had just sold to a sports injury clinic, but they accidentally destroyed five years of clinical and accounting data. Wang was found grossly negligent.Like the Transammonia and Triple Point case, those cases occurred in down economies. Lawyers say litigiousness in the software project world (as in the world at large) is generally inversely proportional to the economy. And so, according to the experience of seven legal specialists interviewed for this article, the number of software disputes coinciding with the recent recession point to more IT project court battles: "[Disputes] sprung back up aggressively in 2002, 2003.""Our firm has seen about a 300 percent increase in cases in the last five years.""The economy went down.... People started suing each other.""As an arbitrator in these cases, I\u2019ve seen a significant rise in disputes in the last three to four years.""We\u2019re getting more and more work as honest brokers [or mediators] in cases just like this.""There\u2019s no doubt there\u2019s a rising tide of cases."What\u2019s different about software project disputes than other corporate lawsuits and what most shocks lawyers about this particular cottage industry is the utter avoidability of many, if not most, of the complaints. These suits are not about failure of projects so much as they are about the failure of CIOs and their executive brethren to legally prepare for failed projects. When they fail to prepare, CIOs can expect to end up in one of three types of dramas: Mediation is the least nasty and most likely to save the project, which it did for Sheleen Quish, vice president for marketing and global CIO of U.S. Can. "It\u2019s like a marriage," she says. "You go to a counselor first, not divorce court."Arbitration is a nastier genre where witnesses and evidence are used and in which there are losers and ostensible winners. Litigation is the very center of software project dispute hell. Phrases you can expect to hear attached to this kind of action are "years of depositions and discovery" and "millions of dollars." (See "Dispute Resolution Pain Index," opposite page.)In any of the three cases, these are mainly contract disputes. And while anticipating disagreements is one of the primary functions of a contract, the written agreements for software projects don\u2019t seem to do that at all, says Hillard Sterling, a litigator and partner at Much, Shelist, Freed, Denenberg, Ament & Rubenstein in Chicago. In any other industry, many of these cases would never become cases because both sides would have signed a contract that accounted for what lawyers would consider easily foreseeable problems.But contracts executed on software projects are often woefully inadequate, according to lawyers like Lee Gesmer. "I\u2019m astounded by some of the [contracts]," says Gesmer, who is a partner at Lucash, Gesmer & Updegrove in Boston. He has also served as mediator and arbitrator in software disputes. He says he\u2019s seen contracts signed by multibillion-dollar companies that are so one-sided in the vendor\u2019s favor that "it appears the customer had no legal involvement in the contract. I walk around the office showing it to other lawyers saying, Look at this," says Gesmer.In many cases, adding just a few words to a contract to specify who\u2019s responsible for what and when would prevent multimillion-dollar disputes (see "A Crash Course on Avoiding Software Project Disputes," Page 45). True story: "Acme," a Fortune 500, deployed a system in which response time to database queries was a joke?a minute or more with several concurrent users. Simply unusable, according to a lawyer involved in the case. Acme told the vendor to fix it; the vendor told Acme to upgrade its hardware, which would cost Acme $1 million. Acme felt the vendor misrepresented its systems capabilities and that the vendor should pay for the upgrade. Acme sued.The case is pending, but who wins is beside the point. Acme could have avoided all of this if its original contract had contained a simple performance clause: one paragraph that specified acceptable query response times and what hardware the vendor would be required to supply to make them possible.Get Comfortable With Lawyers In The Room Lawyers lay much of the blame for all of this at the feet of the CIO, whom they call an "underachieving, underinvolved" participant in software project contracts. "From what I\u2019ve seen, it\u2019s uncommon for an executive team, including the CIO, to have a solid game plan going into these projects," says Sterling, who started a full-time technology litigation practice a year ago after handling such cases for five years.The lawyers attribute this lack of planning to three factors. One, CIOs are inexperienced with contract law and either fear demonstrating that inexperience to other executives (each project is, in large part, the CIO\u2019s baby). Or as Gesmer says, "They\u2019d rather delegate that to someone else." Sterling says the IT executive\u2019s ethic, born out of years of hustling his way into the executive circle by proving his worth, is to demonstrate accomplishment?not to avoid dispute. Being the one to prove you mitigated a potential lawsuit, he says, is difficult and often disregarded anyway.The second factor that keeps CIOs from playing the key role in brokering good contracts is the emotional optimism that always accompanies the so-called BTP, business transformation project. Love and romance, says Bill Bierce of Bierce & Kenerson in New York City, probably explains why a billion-dollar company such as Transammonia, which had the resources to blanket the project with due diligence, didn\u2019t do more of it. When you\u2019re about to get married, no one wants to be the one to bring up prenups. "The human problem is, you want to trust each other, you really do," says Bierce.The third factor, and the most self-serving for the lawyers, is that CIOs don\u2019t cultivate legal experts outside of their own corporate lawyers, who themselves lack experience with software contracts. Take Charlie Talmadge\u2019s experience as a lesson. Talmadge, the former CEO of Merchants & Businessmen\u2019s Mutual Insurance, experienced a software dispute that eventually went through arbitration."I always say this now," says Talmadge, now retired. "If we would have hired expert lawyers up front and spent $20,000 having them argue with the software vendor as we were putting together the deal?instead of relying on our own lawyer, from Harvard, who also didn\u2019t know anything about software?we would have never signed that contract. We would have been better off."Winners In Court, Losers At The Bank Abraham Lincoln once said that in court, "the nominal winner is often a real loser?in fees, expenses and waste of time." Talmadge knows this firsthand. "I took over Merchants & Businessmen\u2019s, and we were trying to install this new commercial policy software system that was supposed to be The Answer. It wasn\u2019t. It wasn\u2019t anything we thought it would be. It was beyond the pale. So the vendor [now out of business] says, \u2019Look we know we screwed the pooch on this, we\u2019ll fix it. Just give us a quarter-million dollars.\u2019 I said, \u2019That\u2019s easy: No. Make it work.\u2019 The president of the software company is in my office and asks me to think it over. I say, \u2019Do it or we\u2019ll sue you.\u2019 He said, \u2019The only people who will win then are the lawyers.\u2019 That was the only thing he ever said that was right." He laughs."We went to arbitration. We had switched CIOs in the meantime [for unrelated reasons]. The old one came back as a witness, and he had documented the nonsense very well. We figured we\u2019d kick the crap out of them. After three or four weeks, I\u2019m paying lawyers and expert witnesses about 400 bucks an hour. It turns out we didn\u2019t have strong enough language in our contract?and we totally missed the [terms contained in the] delivery and acceptance clause?it was that simple."Talmadge says the case resulted in a $250,000 settlement. Legal fees came to $150,000 of that. So Merchants took back $100,000, which was equal to its license costs plus $50,000."The worst thing in the world is being in litigation when you\u2019re right," Talmadge continues. "Because you\u2019re spending money to get what you\u2019re supposed to have anyway."CIOs must work with lawyers to prepare contracts for the eventuality of a dispute, because often the vendor sues the buyer first, just as Triple Point sued Transammonia before Transammonia countersued. (In nearly all cases, experienced litigators say, lawsuits beget countersuits.)But even trial lawyers will tell a CIO that litigation, especially over a software project failure, is a good card to hold but almost always a bad one to play.First, there\u2019s the matter of timing. When projects start to slip, "the CIO has a vested interest never to sue the vendor, because it will acknowledge his failure," says Tobey B. Marzouk of Marzouk & Parry, an arbitrator who has handled more than 50 software project disputes. "Then, at some point, I can\u2019t say precisely when, the CIO has a vested interest in suing the vendor, because it will prove the vendor\u2019s failure" and lead to recovery of lost money. But timing this right is tricky at best, Marzouk adds.Even in cases where the timing seems right and the cost-benefit equation clearly favors court action, there are wild risks. Marzouk tried a case with $13 million at stake in which, he says, his client had a strong case. But they lost. Not only did the company fail to recover its $13 million, but it also had to pay a $1.5 million counterclaim. The High Cost Of Gathering Evidence Professor Thomas Richards of the University of North Texas, who cowrote a paper on software failure litigation (see "Information Technology Litigation and Software Failure" at www.cio.com\/printlinks), doesn\u2019t paint a pleasant picture, either. Richards used to serve as an expert witness in software disputes. "I gave it up," he says. "I said to heck with this. It was too stressful. I remember I\u2019m sitting in this room with four lawyers watching a screen with this [software] application on it. I was the one that was supposed to extract information from it. They\u2019re there charging their clients $300 an hour, and I\u2019m charging them $250 an hour. And after all that, the case came down to one memo out of 10,000 that showed the vendor was using its client to basically beta-test the software."The process of recording witness testimony and gathering evidence is "a black hole of costs," says Sterling. "It can take years with thousands of documents and dozens of witnesses." For those reasons, arbitration has emerged as a common alternative to court proceedings?so much so that arbitration clauses have become common in well-written software project contracts.It takes arbitrators less time to wend through evidence and arrive at a ruling than court?weeks or months rather than years (in some cases, arbitration can be expedited). The shorter process means it costs less than court (though arbitrators can make $3,000 per day). It\u2019s private. And, know that if you\u2019re the one with the beef, many lawyers we talked to say arbitration tends to be a slightly pro-plaintiff forum?just as a baseball player who goes to salary arbitration can expect a healthy raise.But not every distinction between arbitration and court favors the former. For example, the rules of evidence are much looser in arbitration. "I had a case where several other companies had the same problem with a vendor that my client did," says Marzouk, "and we corresponded about it. They told me they couldn\u2019t testify because of confidentiality agreements. So I submitted my correspondence with them as evidence. It was devastating even though it was hearsay. It never would have gotten into court." (Of course, Marzouk adds, the opposing side can return with hearsay in kind.)If full-blown litigation is hell, arbitration is somewhere just south of purgatory. Asked to characterize arbitration, Gesmer rattles off: "Unpleasant. Distasteful. Antagonistic. Extremely unpredictable." In other words, arbitration is, well, arbitrary. And you spend time and money there that should be going into running the business."What I learned is, if you\u2019re in arbitration, you\u2019ve lost," says Talmadge. "You\u2019re sitting around a table with the president, the CIO, maybe a director and a lawyer? It\u2019s already over."Before Losing Control, MediateU.S. Can\u2019s Quish saw her warehouse management software project failing. The first two deployments (of 10) went disastrously bad. Both sides had underestimated the job. The overall project plan was slipping. Tension was high. The monthly executive steering committee meetings were, in Quish\u2019s words, "very unpleasant." And it was in this hostile environment, after two years of working with this vendor and with eight implementations to go, Quish says she and the vendor account rep decided jointly "to do something about it." He suggested going offsite. She volunteered to facilitate.So in this case, Quish herself was the mediator. The offsite meeting started with dinner, where work-talk was expressly verboten. "Two years working together and we had never relaxed," she says.The next morning she started the mediation with a simple question: "What are the issues that prevent us from being successful? No holds barred, put it on the table. It was an encounter. An intervention."Both sides let it fly. And it worked. They figured out what was wrong with the project and how to make it right.If you want to save a foundering project by legal means, mediation is probably your only hope, and even then, it\u2019s far from a guarantee. Among the three dispute resolution measures, it is the least antagonistic and the least formal. There are no witnesses or evidence. Most of the time it\u2019s voluntary.Most of the time, too, the mediator is an independent third party. Often, it\u2019s a retired judge like William McDonald. "I like to start with a joint session, with both sides\u2019 principals, including the CIO," says McDonald. "I let the lawyers talk, then I try to get the principals to state their cases without the legalese. The reason is they\u2019ve probably been getting information tailored to what they want to hear until this point."After this initial meeting, McDonald says, the parties often retreat to separate conference rooms, the mediator shuffling between them like a parent brokering peace between bickering siblings. "At some point, as I sense we\u2019re getting closer to a compromise, I\u2019ll get more aggressive and evaluative. I\u2019ll start saying \u2019They\u2019re going to say this in court. How would you combat that?\u2019 And I\u2019ll start suggesting remedies, specific provisions, which don\u2019t involve arbitration or litigation."McDonald says mediation typically lasts a few days, though the days can run well past midnight.Good contracts will specify the rules of engagement with mediation, including when a mediator should be called in, how the process will be structured, and, in many cases, who will mediate. Mediation hurts less than arbitration and litigation. But it\u2019s relatively ineffective in troubled projects. The worse off a project, the further divided the sides, the less likely mediation will help. In the Transammonia and Triple Point case, mediation might have proven effective in August 2000, when PeopleSoft and Triple Point were lagging with the interfaces but still had months before their deadline. Six months later, with stop-payments flying around, mediation would have been useless.At Quish\u2019s offsite meeting, the sides worked out their differences. The next eight implementations went smoothly; U.S. Can finished them in the space of a year, or roughly the same time it took to get the first two done. "Before you start, inject the idea that if and when we hit roadblocks, we will invest in some sort of third-party mediation," Quish says. "Actually, I loved doing it. You see such an improvement just by going through the process. If we hadn\u2019t done it, the project would have been even more strained. No one would have wanted it, but it would have ended up in litigation."To McDonald, mediation is control. "In arbitration and a courtroom," he says, "someone wins and someone loses, and often the dispute becomes public. In mediation, the parties haven\u2019t lost control yet. It\u2019s amazing how often parties discover their common interests in compromising just by going through the process. I had one case where we had an agreement in principle at 7 p.m. The lawyers for each side then went into a room and bickered over the terms until about 10 p.m."At the same time," McDonald continues, "the principals were in another room, working out their next deal."The Cio In The Witness ChairIt\u2019s quite a battle up here," says Sterling, who talks on a cell phone during a recess from IDX Systems Corp. v. St. John Health System, in U.S. District Court in Detroit, Judge Denise Page Hood presiding.Sterling is colead counsel for IDX, a software developer, which is suing St. John Health for, according to the court documents, allegedly backing out of a project. It would have been a fully integrated clinical information system that would tie together seven hospitals and more than 60 facilities, but St. John allegedly was short on money, and IDX is claiming breach of contract. St. John has, of course, countersued. (The parties declined to comment about their lawsuits.)The issues in this case aren\u2019t much different from the Transammonia and Triple Point case, except there\u2019s no subcontractor, and this case is taking longer. The suit was filed in April 2000. Many millions of dollars, Sterling says, rest on "two or three sentences in the contract and a couple of memos."Overall, the trial phase is expected to last four to six weeks, or as Sterling says his clients think of it, "30 percent of the fiscal quarter."Soon he\u2019ll be going back into court to question one of his star witnesses. She\u2019s in her second day of testimony, and Sterling says she\u2019ll testify again tomorrow and maybe two more days after that. In the past three years, she\u2019s been deposed three times.She works for St. John Health System, the company his client is suing. Her name is Claudia Allen. She\u2019s the CIO. Postscript: Before this story went to press, IDX v. St. John was settled under confidential terms. Sterling says the judge pressed for a resolution.Not only will judges push for settlement ?even 21 court days into a trial?"they\u2019ll elbow you in the ribs if they see any willingness at all to settle," Sterling adds. In this case, the judge\u2019s rejection of several requests by the defense to limit damages motivated the sides to settle. "I will declare victory," Sterling says.As for St. John\u2019s CIO, Sterling says he was able to contradict her testimony during his examination using document exhibits. "It was a dream for litigators but four days of hell for her," he says. "She was exhausted. If you want CIOs to know the perils of getting into this, I wish we could just show them a videotape of that testimony."