The Zombie
Detector.
Is your project alive or just twitching?
A Zombie Project consumes resources but creates no value. It looks alive on dashboards but is dead in the marketplace. These walking dead are draining your organization's capacity to innovate.
Select all symptoms that apply to your project.
The Autopsy Checklist
Started because a competitor announced something similar—not because you identified a real problem to solve.
Exists because a senior leader championed it—and nobody wants to tell them it's failing.
Success is measured by activity (meetings held, code committed) rather than business outcomes. Nobody can define what "done" looks like.
The definition of success has changed three or more times. Each time it gets vaguer.
IT is building it alone. The people who will actually use this haven't been consulted in months.
The original sponsor has left, retired, or moved on. Nobody claims ownership anymore.
"We just need 6 more months" has been said more than twice. The deadline has been pushed back repeatedly.
Good people have quietly transferred off the project. The team is now staffed with whoever was available.
No actual customer or end-user is asking for this. It solves a theoretical problem that nobody experiences.
The business case depends on adoption that has no evidence. "Users will love it once they see it."
It requires integrating with systems that don't want to be integrated. Data architecture is an afterthought.
The tech stack was chosen because it's "cutting edge" rather than because it fits the problem.
Diagnosis
0/12Kill It. Learn. Move On.
Stopping a bad project is as important as starting a good one. A killed zombie releases resources for experiments that might actually work.
Move funds to safe-to-fail experiments with clear hypotheses.
A killed project with documented lessons is an asset, not a failure.
"If we deleted this today, who would complain?" If the answer is "nobody," you have your answer.
Publicly recognize the courage to stop. This builds psychological safety for future experiments.