You may have just inked a fabulous outsourcing deal. But as the CTO of the Red Cross has learned, you have to be prepared for the endgame.

The last thing most CIOs want to think about as they embark on a major outsourcing deal is what will happen when it ends. After all, reaching the decision to outsource some or all of an organization's IT services requires a great deal of business case analysis, transition planning and soul-searching about the best way to handle the many human issues involved. Assuming that the organization views the pending transaction as positive, or at least acceptable, this is a time of high excitement and tension, and it may not seem like the ideal time to plan how to get out of this arrangement in a few years. But you ignore the endgame at your peril. Sourcing deals are risky, and some of the biggest risks are not visible in the rose-coloured days leading up to deal consummation.

Fortunately, these risks will come to light not at the very end of the deal, but somewhere midstream, as they did a few years ago for me and my colleagues at General Motors. GM manages one of the industry's largest IT outsourcing agreements with EDS. Many factors make this arrangement unique, including the fact that it was created in 1996 when GM divested EDS with the proviso that EDS would maintain a more or less exclusive lock on GM's IT business for the next decade. Ralph Szygenda was hired as CIO for GM's newly formed Information Systems & Services organization, and he hired a couple of hundred IT executives and managers, including yours truly, to drive GM's technology strategy and manage the EDS outsourcing deal. I led three organizations in my time there, including the operations portion of GM North America's multibillion-dollar agreement with EDS, which is set to expire in 2006. In 2001 (the midpoint of the contract), we began to analyse the opportunities and risks involved in the expiration of that arrangement. While that story has yet to play out, the risks we identified are relevant to any sourcing deal, as I have since discovered in other contracts I have worked on, including here at the American Red Cross.

The first lesson from my GM days is that long-term exclusive outsourcing isolates the organization from the market. You still have widespread access, of course, to technologies and solutions, but you have granted one vendor the exclusive right to understand your business in depth. The more this outsourcer learns about you, the harder it will be to change your sourcing choice once the contract is up. This dynamic exists not only because this vendor understands your business well, but because others have not had the opportunity to do so. There is simply no knowledge base in the vendor community about your needs and, therefore, only limited ability to respond to downstream sourcing opportunities.

The way to mitigate this risk is to create competitive events in your sourcing arrangement. Even in its exclusive deal with EDS, GM had the ability to competitively bid some services - long-distance telephony and voice mail, for instance. We also rebid all of the IT services in GM's locomotive division. That gave other vendors the opportunity to learn about GM's needs. Many contracts contain provisions to rebid some services to keep pricing and terms competitive, but the chief benefit of these competitive events is to educate other vendors about your business. Keep such competition reasonably modest in scope so that it doesn't undermine your deal, but use it strategically to build competence in the vendor community.