Two thirds

Whenever we break down work, (e.g. sessions in an Open Space, continual improvement backlog, teams in a hackathon, experiments) here’s my rule of thumb for measuring success;
For each iteration, half is OK, two thirds is good, three quarters is great.

Two thirds of the people in each task engaged.
Two thirds of the teams give a useful result.
The first time I’ll be happy with half. After five iterations I would expect much better. England’s Law: it takes about three iterations (plus or minus) to get any good at anything.

It’s not an army parade. It doesn’t need to be even close to perfect to be a win.

The other third is *normal* failure. It’s fine. It has value as learning for next time.