War-game a growth experiment before paid channels, product flags, and support queues absorb a bet that only looked clean in a slide.
Growth experiments fail when the success metric is a vanity rate, when the holdout is too small to read, when support and sales never see the offer language, and when a "temporary" discount becomes the permanent price. A colorful funnel chart is not evidence that the experiment survives scrutiny.
Freeze the bet
State the hypothesis in one sentence, the primary metric, the guardrails you will not violate, the stop criteria, and the budget and calendar window. Attach the creative or offer copy, the audience definition, the holdout design, and the exception policy for sales. If growth, product, and finance disagree on what "win" means, split the package until one move is under fire.
Write the decision you will make if the war game finds nothing new, and the kill criteria if CAC, churn, or ticket volume crosses a named line.
Seats that matter
- Finance. Unit economics under realistic conversion and refund rates, not the optimistic slide.
- Product and eng. Whether the flag, landing page, or pack change can ship without silent dependency debt.
- Support and CS. The ticket and churn stories the offer will create in week one.
- Competitor. How the public offer will be quoted against you.
- Skeptic. The claim that looks strongest and is least sourced.
Give every seat the same pack. Secret "we will fix quality later" notes for growth only create fake agreement.
Loop the review
Feed Pingpong the experiment brief, creatives, and metric plan. First pass steelmans the bet. Later passes attack from finance, product, support, and competitor seats. Final pass turns surviving objections into a tighter holdout, clearer stop rules, or a delayed launch. Delete dual-counted revenue and invented organic lifts.
Force a day-after narrative: what happens if the offer leaks to the full customer base, if a journalist screenshots the landing page, or if support cannot explain the fine print. If those stories are stronger than your guardrails, fix the package before spend. Separate a true experiment from a permanent pricing or packaging change. If the memo blurs them, run the packaging or pricing war game as its own move.
Related: war-game a pricing experiment, stress-test a marketing campaign, stress-test a feature flag rollout, war-game a churn response, and the war-game decisions hub. Process: how to run a Pingpong.