A weak experiment starts with “let’s try a green button”; a strong one starts with a hypothesis you can prove wrong. This tool assembles A/B test hypotheses in the industry-standard If/Then/Because structure so every test you plan names a change, a measurable outcome, and a reason rooted in user behaviour.
How it works
Each hypothesis is built from three slots. The If clause names a concrete change to a real funnel element (headline, CTA copy, form length, social proof, page speed). The Then clause names a measurable outcome and a direction, such as increase signup completion rate or reduce checkout abandonment. The Because clause states a behavioural rationale, for example because reducing fields lowers cognitive load. The generator draws each slot from a focus-matched pool, so a checkout hypothesis pulls checkout-relevant changes and metrics rather than random ones.
Tips and example
A finished example looks like this:
If we shorten the signup form from 6 fields to 3, then signup
completion rate will increase, because reducing the number of
required fields lowers friction and cognitive load.- Test one change per hypothesis. If you alter the headline and the button colour together, a win tells you nothing about which one worked.
- Always attach a real baseline and a target. “Increase conversion” becomes testable as “increase from 4.2% to at least 5.0%”.
- Prioritise hypotheses by expected impact and ease, not by how clever the idea sounds. The boring form-shortening test often beats the redesign.