The pilot project goes well beyond designing a simple prototype. Its purpose is more to identity the set of tasks needed to perform the full migration and to mitigate the main technical uncertainties.

The greatest risk is defining a scope that is inappropriate. The scope should be adapted to the company's priorities. The classical mistake is to aim for completeness when selecting the features to test. Choosing appropriate services and identifying a representative sample of the user population are the keys to a successful pilot.

The final evaluation of the pilot should clearly distinguish between technical issues on one hand and the end-user evaluations of the tools on the other.

The pilot phase is a good opportunity to evaluate the amount of support that users really require. When appropriate, ambassadors can replace or complement the traditional helpdesk.1 TCO: Total Cost of Ownership

2 ROI: Return On Investment