Case studies tell us about successes more often than failures. This is because parties involved in a failure don’t like to provide information, or even admit that a project was a failure. Yet, we can find out a lot more from failure than success. Consider an engineer who builds a bridge. The 10 well designed bridges that stand for 100 years tell us less about engineering than the one bridge that falls down. Was it metal fatigue, poorly manufactured parts, why didn’t an inspector spot the problem, were there clues that were dismissed, did we learn something new about wind forces, or were promising but untested material used? Success may only tell us that we didn’t fail. Failure, when carefully studied, tells us what went wrong.
Over the past few years, there has been a massive wave of outsourcing. When outsourcing is done in a hurry or under duress, some projects will be executed upon poorly or inappropriately. The collapse of global financial markets fueled many projects, and has provided us with a massive test environment to find out what works and what doesn’t. The data from many failed experiments can be found in our court system, where former business partners are arguing over who was at fault for the failure. An excellent example of this is the $1.4 billion contract between the State of Indiana and IBM.
Indiana wanted to outsource its welfare processing systems. Outsiders rarely learn the details of failed outsourcing projects. The people involved either won’t talk about it or won’t admit the project failed. In this case, we know the details because the client and the vendor chose to sue each other, and the details of their disagreement are now in the public record. In the words of presiding Judge David Dreyer, "Neither party deserves to win this case. This story represents a 'perfect storm' of misguided government policy and overzealous corporate ambition. Overall, both parties are to blame, and Indiana's taxpayers are left as apparent losers."
There are many lessons to be learned, but here are the top 8:
1. Change Requires Commitment: This contract was to transformation, “the nation’s worst welfare system (criminal fraud, rampant incompetence, favoritism, etc.),” introduce a new service delivery model (dial into rotating case workers vs. travel to an office and work with one case worker… for life), reduce costs, and fix decades of federal regulation violations. Accomplishing any one of these goals would have been remarkable. Achieving them all would be unprecedented. Obviously, the vendor is going to need absolute and unconditional support from the client. Instead, court evidence shows that state representatives intentionally undermined the program (and violated the contract) by interfering with IBM’s management of (politically connected) sub-contractors. If clients cannot lead and support change…change cannot happen.
2. Learn From Other Examples: As part of the preparation for the contract, similar programs in Texas and Florida were examined. These programs failed (or were failing), in very much the same way that Indiana's eventually would. The problems in Texas were so severe that, “the rollout of the project was stopped.” However, many of the same mistakes were made. IBM decided that these issues didn’t apply to them, or that they would be able to manage the issues. IBM is unquestionably a great company, but this contract needed was more than IBM could provide. Mega-contracts blind vendors to flaws and limits; the client must see physical evidence that risks have been sufficiently mitigated.
3. Mega Contracts = Mega Risk: A single big contract is riskier than several smaller contracts. You may choose to take the risk because a single big contract can cost less to manage. If it is successful. If not, a failed big contract is very, very expensive. The potential benefits of lower management cost can be compared to the increased risks of failure. And the cost of risk mitigation can also be factored in. Court records show acknowledgement of some risks, but there is no evidence that the client or the vendor ever did the math and calculated the costs. Size alone is a factor for failure, but size tends to have co-factors (as this contract did) that lead to failure. High risk contracts need to have in depth risk analysis and mitigation.
4. Change Happens: Many outsourcing contracts are designed to drive change. However, outsourcing contracts that are supposed to drive change, but do not allow the vendor to make changes, tend to fail. In this case, the client tightly controlled the change mechanism and did not approve most vendor requested changed. The conditions of the program changed: new programs were added, the volume of work expanded, etc. Yet, even for client initiated changed and expansions, they did not allow the vendor to add staff (and cost) or make other changes. This was a 10-year contract. Over a decade, things will change in ways you cannot expect. For example, the economic downturn doubled the volume of requests for welfare aid. Just saying, “we don’t want changes” is not a change management plan; if you want a successful program, you need a reasonable mechanism to approve and implement change.
5. Disputes Lead to Lawsuits: Nobody wants to go to court. However, if neither side of a dispute is willing to work out the issues, you’re headed to court. A smaller vendor might hesitate to sue the government, or might give in when threatened with a lawsuit, but huge vendors like IBM have equally huge legal departments (another risk of mega contracts). Everyone has disputes, but when communication stops then other avenues for resolution are closed, and both parties start to think about lawsuits. An old rule of project management is, “Settling a dispute in court is the most costly and least effective solution.” When communication starts to close down, do everything to keep those communication channels open; make compromises and be inventive now, because a court-ordered solution will be more expensive.
6. Be Consistent: In the first three years Indiana officials repeatedly agreed the program had succeeded, and (as per the contract) told IBM to move on to the next stage in the program. When the State of Indiana sued IBM, they said the program had failed, and had been in failure for years. This sort of inconsistency not only undermines your credibility in court, it undermines your credibility anywhere. You have a right (and duty) to change your position when new evidence becomes available; however, if you add unsupportable embellishments, you will do more to undermine your credibility that to support your argument.
7. “Prefect Execution” Doesn't Exist: Clients often build massive contracts as insurance that vendors will understand what they want and execute the contract flawlessly. In real life: assumptions are wrong, conditions change and the goalposts you’re aiming for move. Still, both the client and the vendor will select the individual clauses that support their position. The courts take a different position; a judge is not interested defining perfection; judges are interested in defining what is reasonable. Unless one party or the other was completely incompetent or malicious, a judge will seek a compromise position that will make neither party completely happy. Going to court doesn't increase your control, rather it sharply reduces the control of both parties.
8. Both Sides CAN Lose: This is the culmination of all the other lessons, and perhaps the most important. As the judge put it, all three parties lost: the State of Indiana, IBM and the state’s taxpayers. Each of these issues was avoidable, but each problem led to the next until the chain of events was too strong to break. Every person who ever ended up in court asks themselves, “When did this go wrong.” And the answer is always, “Long before the lawsuit began.” Difficult problem can be overcome, but not without effort and planning. Problems must be identified and resolved when the client and vendor BEGIN to pursue a different agenda; if you wait too long the momentum of events will reach a point where the matter is past resolution.
In the end, this was all about common sense. All parties appeared to be intelligent and resourceful, and were more than sufficiently competent to know that significant problems needed to be overcome. Yet, the court found that the real issues… self-interest, conflicting agendas, lack of competence, and unacknowledged risk… were largely ignored. Until it was too late. While they didn’t learn, we can. Whether you are planning a mega deal or something more modest, make sure that you don’t repeat the same mistakes!