In every class I teach, pilot projects come up for discussion, to which I respond: pilots are good. It's a learning experience, and it's a way to work out the message when you take it mainstream.
- Pick a project that is mid-size for your enterprise re duration and budget
- Be sure to pick a project that has some stakeholder support
- Pick a project that is obviously important--but not too important
- Don't wade into a "wicked" social policy problem and offer a quick fix on a small piece of the puzzle
- Don't pick a project that is too fuzzy on 'done', competing too much with on-going operations, and thereby muddying the metrics
- Don't pick a project that is "bet the business" for either a stakeholder or the enterprise. If you're going to gore someone's ox, pick a small ox.
- Pick something you can actually get done
- Pick something that you can actually measure the difference with some benchmarks of other methods
- Pick something that functional managers are like to cooperate with (not too many functional nemisis while you work out the method kinks)
- Pick something that will exercise the methodology you want to adopt. No point in piloting only one or two points of a new method
Bookmark this on Delicious 
