The 10-to-1 Input/Output Kill Test
When you pour 10 units of effort in for 1 unit of output, the project has run its course.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 4
- Confidence
- 88%
A framework for deciding when to shut down a beloved but stalling project. The core signal is a lopsided input-to-output ratio: when the team pours enormous effort into launches they're proud of and the metrics barely move, the energy isn't present in the system. Combined with a 'you know it when you see it' read on traction and a rising opportunity cost, this tells you to call it — cleanly and early — rather than protracting the decline.
Origin
Mike Krieger explaining his decision to shut down Artifact, his post-Instagram news startup, and the peace he made with calling it when he saw the signals.
Core principles
- 01Set the success bar honestly up front (not everything can be Instagram-scale)
- 02The kill signal is a lopsided input-to-output ratio, not a single bad metric
- 03Traction is 'know it when you see it' — look for legs that compound on their own
- 04Rising opportunity cost against a bigger wave raises the bar to keep going
- 05Calling it early and cleanly earns respect and frees people for better work
How to run it
- 1
Set an honest success bar upfront
Define what success means for this specific project rather than defaulting to an impossible benchmark like Instagram DAUs.
Pro tip Very few products reach mass-consumer scale — pick a bar the project could realistically clear.
Watch out An unrealistic bar makes every project look like a failure and clouds the kill decision.
- 2
Watch the input-to-output ratio
Track how much blood, sweat, and effort goes in versus how much the metrics actually move. A ~10:1 input-to-output ratio signals the energy isn't in the system.
Pro tip You're looking for something that 'has its own legs under it' and can compound, not effort-forced growth.
Watch out Being proud of a launch is not traction — pride and moving metrics are different signals.
- 3
Weigh the rising opportunity cost
Assess what larger wave you'd be forgoing by continuing. When a bigger shift (e.g. AI) is changing everything, the cost of staying rises.
Watch out Structural headwinds you don't control (e.g. deteriorating mobile web) can cap upside no matter the effort.
- 4
Make the call cleanly, not protracted
Once the signals converge, decide and communicate rather than dragging it out for another 6–24 months hoping it turns.
Pro tip Founders respected Krieger for 'calling it when you saw it' — and some killed their own doomed projects sooner after seeing it.
Watch out Expect an ego bruise; separate your self-worth from the last project's outcome.
In the wild
Krieger's team put blood, sweat, and tears into launches they were proud of, but metrics barely moved — roughly 10 units of input for 1 unit of output. Facing deteriorating mobile web, weak word-of-mouth spread, remote-team friction, and the rising opportunity cost of AI, they decided over a couple of months to shut it down.
→ He was 'really at peace' with the decision; the reception was overwhelmingly supportive, and freeing people to work on more interesting things became a good legacy for Artifact.
Common mistakes
Protracting a doomed project for another year or two
Dragging it out only to fundraise and discover the same lack of traction wastes the team's time and compounds opportunity cost — call it when the ratio is clearly inverted.
Judging against an impossible benchmark
Measuring a niche news app against Instagram-scale DAUs distorts the read; set a realistic bar so the kill/keep signal is legible.
Is it for you?
Best for
Founders and product leaders deciding whether to keep pushing or shut down a stalling project they care about.
Not ideal for
Early projects still in genuine iteration where you haven't yet given the core bet a fair test.
From the transcript
“Like the way I put it, it's like 10 units of input in for one unit of output versus the other way around”
“like metrics would barely move. I'm like there the energy is not present in this product in this system”
“you kind of know it when you see it in terms of like user growth and traction”
“kudos for calling it when you saw it and not like kind of protracted”
From the episode
Anthropic’s CPO on what comes next
Mike Krieger (co-founder of Instagram)