Kill Criteria

What is Kill Criteria?

Kill criteria are conditions agreed before an initiative, product or experiment starts that will trigger stopping it, such as missing an adoption target by a set date or a key assumption proving false. Setting them in advance makes it easier to stop work that is not succeeding, instead of continuing because of effort already spent.

Updated November 2026 · 1 min read


Organizations are good at starting things and bad at stopping them. Once people have invested time and reputation in an initiative, it becomes hard to admit it is not working, a pattern known as the sunk cost fallacy. Kill criteria counter this by deciding, while everyone is still objective, what "not working" will look like.

What good kill criteria look like

  • Specific and measurable, such as fewer than 500 weekly active users after 90 days
  • Tied to the riskiest assumptions behind the initiative
  • Time-bound, with a clear date or milestone for the check
  • Agreed by the people who will make the decision, before work starts
  • Paired with what happens next: stop, pivot or seek more evidence

How companies run it in practice

Kill criteria often appear in business cases, portfolio reviews and stage-gate reviews, and in product experiments as a failure threshold. They work best when stopping is treated as a good outcome that frees capacity for better bets, not as a failure for the team. Some companies hold regular "keep, kill or adjust" reviews of the portfolio to apply them consistently.

Kill criteria vs success metrics

Success metrics define what good looks like. Kill criteria define the point at which continuing is no longer worth it. Most initiatives need both, because performance between the two thresholds calls for judgment.

Frequently asked questions

What are kill criteria?

Kill criteria are conditions set in advance that will trigger stopping an initiative, product or experiment. They are usually measurable and time-bound, such as missing a usage target by a certain date, and help teams stop work that is not succeeding rather than continuing because of effort already invested.

Why are kill criteria important?

They counter the sunk cost fallacy and the reluctance to stop work people care about. By agreeing what failure looks like before starting, teams can make stop decisions objectively and free time and budget for initiatives more likely to succeed.

When should you set kill criteria?

Set them before work begins, when the business case or experiment is approved, and agree them with the people who will make the final decision. Review them at planned checkpoints, such as portfolio reviews or stage gates, and update them only for good reasons that are recorded.