Constraint War Game
Model hard decisions deeply before spending money on a solution.
- Difficulty
- Moderate
- Time to result
- ~days to results
- Steps
- 5
- Confidence
- 95%
The Constraint War Game is a decision process for fundamental business challenges. First define the problem without allowing fresh spending to become the automatic escape route. Generate plausible actions, model what each would require, and discuss their assumptions, consequences, and failure modes until the team can explain why one direction is stronger. The result should become a structural fix, operating process, or clear experiment rather than a one-time patch. David Heath says Bombas uses internal war games and scenario planning because bootstrapped dollars mattered. The constraint trains the team to do the hard analytical work first, while the final commitment prevents scenario planning from becoming indefinite discussion.
Origin
Extracted from The Foundr Podcast
Core principles
- 01A fundamental challenge needs a structural answer, not automatic spending.
- 02Constraints force teams to understand a problem more deeply.
- 03Multiple scenarios expose assumptions before execution begins.
- 04Conviction should follow deliberate examination, not endless debate.
How to run it
- 1
Name the real problem
Describe the fundamental challenge and distinguish it from its visible symptoms. State what must remain true after any temporary resources are gone.
Pro tip Ask what the team would need to change if no new funding were available.
Watch out A vague problem statement produces vague scenarios.
- 2
Constrain the easy answer
Temporarily remove additional spending as the default response. Force the team to examine product, process, structure, and sequencing first.
Pro tip Treat the constraint as an analytical device even when the company can afford to spend.
Watch out Do not reject a genuinely high-return investment merely to perform frugality.
- 3
Build the scenarios
List credible courses of action and model their likely costs, returns, dependencies, and failure modes. Include the option of changing nothing.
Pro tip Assign someone to challenge the favored scenario.
Watch out Optimistic labels are not substitutes for consequences and assumptions.
- 4
Debate to conviction
Talk through each scenario until the team can explain why the selected direction is preferable. Record the assumptions that would cause the decision to change.
Pro tip End the discussion when new rounds stop producing new information.
Watch out Analysis without a stopping rule becomes avoidance.
- 5
Install the answer
Turn the chosen direction into a process, structural change, or bounded test. Review the decision against the recorded assumptions after real evidence arrives.
Pro tip Preserve the model so the next decision can reuse what the team learned.
Watch out A one-off fix does not train the organization to solve the next version of the problem.
In the wild
A subscription founder sees growth slowing and initially proposes more acquisition spending. Using the Constraint War Game, the team models increased ads, a product onboarding repair, and a customer-support intervention. The scenarios show that poor early activation would make extra traffic expensive, so the team tests onboarding first and defines the retention result required before resuming acquisition spend. This is an illustrative application of the Bombas decision process.
→ The team directs effort at the structural weakness before paying to amplify it.
Common mistakes
Starting with the budget
Asking how much to spend before defining the problem makes money the strategy rather than a resource.
Modeling only the favorite
A war game that does not compare real alternatives merely decorates an existing preference.
Debating without deciding
Discussion should create conviction and a testable action, not become an endless search for certainty.
Is it for you?
Best for
Founders facing consequential operating choices where several plausible paths exist and cash could mask weak reasoning.
Not ideal for
Reversible, low-cost experiments where extended discussion would cost more than testing the alternatives.
From the transcript
“We have to be able to solve it without money.”
“you run 17 30 it on a model where you like talk it through”
From the episode
591: From $0 to $3.4 BILLION Selling Socks