Phase 3: Extend
Run a mid-mortem when pilots stall
Pause, look for patterns, and re-decide in the open.
Set the trigger in advance. A workable default: check every pilot at 45 days, and call a mid-mortem when two or three in a row show nothing measurable.
Put it to work
- Pause new scaling. Let running pilots continue.
- Look for repeating barriers: data quality, no protected time, thin trust.
- Sort them: remove now, work around, or change your approach.
- Ask leadership to openly re-decide, and write down what you learned.