Delivery ended at launch
The business case justified starting the work, then disappeared. Delivery was tracked closely until release, and the benefit was assumed to follow. Nobody was asked afterwards whether it had.
Changing the lifecycle
I redesigned the lifecycle so that benefits run through it rather than sitting in front of it. Each piece of work now starts with a stated benefit hypothesis and a named owner. After release, the owner revisits the benefit at set points, and the evidence goes into the portfolio review, where it informs whether related investment should persevere, pivot or stop.
Connecting the measures
Each benefit traces from the product outcome to cost, revenue or risk, so the review knows what it is checking.
Changing what leaders aim for
A new lifecycle changes little if people are still rewarded for the old thing. Product goals had been written as outputs: features shipped, releases made. I replaced them with outcome goals, and that was the larger change.
A capability programme gave product managers the practice to work this way. Coaching helped senior product leaders write goals they could genuinely be held to.
What I learned
The process was the quick part. Changing what leaders felt accountable for took longer, and that is what made it stick.