You have a project that should have died months ago. Everyone knows it. No one wants to be the one to kill it. So it lingers, consuming resources and attention, waiting for permission that never comes.
Write The Death Certificate
One page. What the project was. Why we started it. Why we're stopping. What we learned. This document protects the team and preserves the lessons.
List What's Salvageable
Code that can be reused. Research that still applies. Relationships built. Most projects aren't total losses. Name the value that survives.
Identify The Stakeholders
Who needs to know before the announcement? Who will feel blindsided? Talk to them first. Not to ask permission, but to show respect.
Frame It As Resource Liberation
"We're freeing up this team to work on X" lands better than "we failed." Stopping a bad bet is smart, not shameful. Position it that way.
Set The Shutdown Timeline
Don't let it drag. Two weeks max. Archive the repo. Close the Slack channel. Reassign the people. Clean endings create clean starts.
Hold A Brief Retrospective
30 minutes. What did we learn? What would we do differently? Add the insights to your scar archive. Then move on.
A clean ending instead of a slow fade. Resources freed for better uses. A documented lesson instead of a whispered failure. And proof that your organization can make hard calls.
This works for straightforward kills. When the project has political protection, when egos are attached, when stopping one thing requires renegotiating a web of commitments: that's a different conversation.