Quick vs Full Experiments
Creator plan requiredWhen you create an A/B test from the builder, Flo offers two modes:
| Mode | Best for | What changes |
|---|---|---|
| Quick experiment | Testing one hypothesis at a time | A single dimension (headline, hero, price, CTA, etc.) |
| Full redesign | Bigger layout or multi-block changes | Any blocks and page settings on the variant |
Quick experiments
Quick mode keeps variants aligned except for the dimension you pick. Flo clones your live page, applies the change to variant B, and records which input you changed (changed_inputs) for the flywheel.
Use Quick when you want sequential isolation — run test A, learn, then run test B on the next dimension without mixing variables.
Full redesign
Full mode is the classic split test: edit variant B freely (blocks, cover media, checkout blocks, etc.). Use it when the change spans multiple areas of the page.
Sequential isolation prompt
After you declare a winner on a Quick test, Flo may suggest a follow-up experiment on the next highest-impact dimension. Adopting the suggestion logs the choice and pre-fills the next Quick test — you still start and stop tests yourself.
No auto-promotion
Reaching Significant status does not promote a variant automatically. You review results and Declare winner when ready.