The UGCHooks field guide

Creative testing

Turn exported variants into a test with a question you can answer.

13 practical guides1 topic

Find a guide13 guides shown

Change one creative variable in a hook test

Changing the hook, body, offer, and audience together makes the result difficult to explain.

A hypothesis connects a creative change to an expected observation.

A control is the reference used to interpret new variants.

A useful filename identifies the source and the creative difference.

A matrix helps show which combinations exist and which question each comparison answers.

Platforms can use different thresholds and denominators for similar-sounding metrics.

Hook rate is meaningful only when its view threshold and denominator are stated.

A retention drop near the transition may indicate a mismatch, but the curve alone does not prove the cause.

An opening can attract attention without attracting the right intent.

A handful of events can produce large percentage swings.

A declining result can have causes beyond the creative becoming familiar.

Organic distribution and paid delivery create different comparison conditions.

A retrospective should preserve what was learned, not just announce a winner.