A big bet is hardest to stop once it’s running. Money has been spent and people have defended it in public. Every month there’s a reason to give it a little longer. This Practice decides how the bet ends before it starts. The plan is presented with the condition that would stop it and the way out, and both are approved together.
A tool that supports it
Robust Contingency Architecture is a way of acting, and more than one tool can back it. A direct one is “Kill Criteria,” which the author Annie Duke describes in her book Quit.
A kill criterion is a condition, set in advance, that tells you to stop. Duke’s simple form pairs a state with a date. The state is a result you can check, like a number of customers. The date is the day you check it. If the state isn’t reached by the date, the project ends. No new debate. You write it while you’re calm, because later you’ll be too invested to judge it fairly.
For this Practice, the kill criterion is half of the work. The other half is the exit itself, meaning who makes the call to stop and what in the plan keeps the way out open.
When to use it
Use it for a decision that is expensive and hard to reverse, like entering a new market or signing a long contract. The signal is your own plan with no page about what happens if it fails.
It belongs in the proposal, before approval. Added after the launch, it reads as doubt. Presented with the plan, it shows the risk has been thought through.
How to use it
- Pick the one measure that would tell you the bet is failing.
- Set the state and the date. Write the number and the day you’ll check it.
- Design the exit. Say who stops the project and what happens to the people and the assets. Put a cost on it.
- Build the exit into the plan, with terms that let you leave, like a contract you can end early.
- Present the plan and the exit together, and ask for approval of both. Showing your own plan with its ending attached is the hard part.
Example
The CEO of a regional gym chain wants to open its first location in another state. It will cost $600,000.
She builds the exit first. The state is 800 members, the level where the gym pays for itself. The date is March 31, nine months after opening. To keep the way out open, the lease allows a sublease, and the equipment is rented for the first year.
The board approves both in one vote. On March 31 the gym has 870 members, and nobody has to argue about whether that’s enough.
Phrases you can use
- “Here’s the plan, and here’s what would make us stop.”
- “What result, by what date, tells us this isn’t working?”
- “If we miss that number, who makes the call to end it?”
- “What does it cost us to get out, and can we lower that now?”
- “I’m asking you to approve the exit along with the plan.”
- For a leader, when someone does it: “You showed us how this ends if it goes wrong. What did that change in the contract?”
How to know it worked
The plan and the exit are approved in the same decision. On the date, the team checks the number and acts on it without reopening the debate. Over time, large proposals in the company come with their own stopping point.
In a Skill Deck Track on Decision Making, a person can choose this Practice, and the meetings they already have show when they use it.
skilldeck.ai


