Prepare an A/B Test Plan
Turn a pile of assets or ideas into a clean test plan with one variable per test and clear stopping rules.
The Prompt
Turn these ideas into an A/B test plan: [ideas]. Group them into tests that each change one variable. For each test give the hypothesis, the primary metric, the minimum sample per variant for [baseline conversion rate], and the rule for stopping, continuing, or declaring a winner.
How to use this effectively
- Give your real baseline conversion rate so the sample size estimate is usable, not generic.
- Group ideas so each test changes exactly one variable — combined changes make results unreadable.
- Set the stopping rule before launch so you're not tempted to call a winner early on noise.
Put this prompt to work
Posta connects your marketing context so you don't have to copy-paste parameters.