What I look for in the first week of a stalled program

When I’m asked to look at a program in trouble, I spend the first week listening rather than reading. The documents describe the plan. The people describe what is actually happening, and the gap between the two is usually the diagnosis.

Who can say no

I ask everyone the same question: who can stop this program? If the answers differ, ownership is the problem, not the technology. No amount of delivery effort fixes a program that nobody is allowed to steer.

What the plan assumes

Every plan rests on a few assumptions that are never written down: that a vendor will deliver on time, that a team will be available, that a migration will run cleanly the first time. I list them and test the riskiest first.

A program in trouble almost always knows it. The question is whether anyone has permission to say so.

What the team would cut

I ask the people doing the work what they would stop doing if they could. Their answers are usually the fastest route back to a date the board can believe.