Kill Criteria

Kill criteria are the if-then rules you write on day one that tell the team exactly when to pause or kill the project. They exist because design projects rarely die with fireworks. They bleed out over months of tiny scope additions and missed dates until both sides resent each other and the invoices dry up. The article calls them the section nobody wants to write because it feels like planning to fail. That discomfort is the entire point.

Most briefs are wish lists dressed up as strategy. They list adjectives and attach mood boards. Then week six arrives and the client asks for one more round and nobody has language to push back. Kill criteria fix that by naming the bad outcomes in advance and tying them to automatic actions. You do not negotiate them in the heat of the moment.

This is not a corporate risk register. It is not success criteria with negative wording. Kill criteria are mechanical. Trigger X produces action Y with no committee debate and no vibe checks. Teams confuse them with assumptions or issue logs. Those are passive. These have teeth.

The article lists six that work. If the decider changes without a handoff the new one must re-sign the constraints sheet in five business days or the project pauses. If the budget kill number is reached work stops and you send the real invoice. If two consecutive milestones get missed without explanation the project enters a recovery review. These rules protect both sides.

Linear ships on six-week cycles only because their kill dates have real consequences. Stripe includes docs in the done condition so they cannot get quietly dropped. You can reverse-engineer their briefs from the work that never shipped. The out-of-scope list becomes visible in the features that stayed on the cutting room floor. The same discipline appears in value-based-pricing and change-order language.

Use kill criteria on any project bigger than twenty thousand dollars or longer than six weeks. They earn their keep when multiple stakeholders are involved or when the client is an enterprise used to grinding agencies. Skip them on small friendly logo jobs where the financial downside is low and trust is high. The tradeoff is they can make the kickoff feel slightly tense. That tension is data. Better to surface it early than discover it after you have delivered three extra rounds for free.

Never treat kill criteria as weapons. They must be mutual protection signed by the decider. The studio stays safe from endless unpaid work. The client stays safe from projects that never ship. If you are not willing to enforce them when triggered then do not write them. Empty rules create more problems than none at all.

The one-page constraints sheet plus kill criteria turns the brief from theater into a contract. Without them you are hoping the project stays in bounds instead of designing the bounds. Anti-dashboard thinking applies here too. One clear rule beats twelve ambiguous ones.

Write the exit ramps before any work starts. The brief is the cheapest deliverable in the project and the most expensive one to skip.

A solid kill criteria section turns polite drift into decisive action before anyone loses money or reputation.