A Small Pilot Still Needs an Owner for the Mess

A lesson I keep returning to when judging a “small” product experiment: small scope does not automatically mean low cost. A pilot can involve few users and still create unplanned support, cleanup or follow-up work for someone else.

Before starting, name not only who runs the test, but who handles its failure modes. For example, if a prototype produces a confusing result, who checks it, tells the user, and decides whether to retry or stop?

A useful first trial has a boundary the whole team can see: one task, one owner for exceptions, a review point, and a stop condition. Track the unexpected work too—not just whether the prototype completed the task. If the test succeeds only because someone quietly absorbs the cleanup, that is important evidence about the product, not a footnote.