Hooks and variants
Variants exist so a media buyer has something real to test, and so a win can be explained afterwards.
The short version
Is this for you? For the campaign where one hook is carrying everything and last month's win cannot be repeated, because nobody can say what won. The whole point of a variant set is that the winner is explainable. Not if you are not testing at all; a variant set is only worth building where a buyer will actually put spend behind comparing them.
What leaves your plate. Each version changing one thing and one thing only, angles pulled from research rather than guessed, and the ones likely to lose built cheaply on purpose so the budget proves something instead of decorating it.
What starts working. A set your buyer can actually read, and a next round that begins from a result somebody can point at rather than a hunch everybody shares.
The smallest way in. A free audit: three fields, and a written reading back, usually within about three working days. It commits you to nothing.
What to read next. The five levels to see where this fits, or the casework to see it done.
A set built so a win can be explained afterwards
Variants isolate one variable
Three versions that differ in five ways teach nothing, because a winner cannot be explained and therefore cannot be repeated. Ours change one thing at a time: the hook, or the angle, or the proof.
Angles are chosen from research
The angle is chosen from the audience's pain points and awareness stage, which is work that happens before anyone opens an editor. That research runs inside every engagement here and is handed to the partner at the end of it.
Most of a set is built to lose, so it is built cheaply
A variant set only works if the versions that fail were cheap to make. They come out of one shoot, one master and one grade, with the difference living in the first three seconds and nowhere else. Finishing all six of them to the same standard would be a beautiful way to spend a budget on discovering which one people wanted.
Built so optimisation has somewhere to go
A set that covers hooks and angles gives a buyer a real matrix to run. When one cell wins, the reason is legible, and the next set starts from a fact instead of another opinion.
The order the variants run in
What a set usually contains
Several hooks against one body, then the winning hook against several angles, then the winning pair against different proof. Sequential rather than simultaneous, because a budget spent on twelve unrelated variants at once learns less than the same budget spent in three rounds.
The hook is chosen on evidence
Retention data where it exists, and the research where it does not. Never on which one felt best in the edit, which is the most expensive way anyone has found to pick a hook.
One version is not a creative decision. It is an untested assumption, produced to a high standard.
A set that isolates one variable is the only kind whose winner can be repeated.
What changes for you
A buyer with a set they can actually learn from, and a next round that starts from a result somebody can point at.
Related: Attention and eye path · Market and audience research · Ad creative · A/B testing
Send what is running and we will read it as an ad account, not a showreel: angles, variants and where spend is leaking. The reading comes back in writing.
Request your free audit.
Three fields, and a written reading back: a fit snapshot, two or three prioritised opportunities, and a recommended next step.
A person reads the channel and writes the audit by hand: a considered read typically takes three working days. That is the usual shape, not a promised turnaround. We use these details only to reply to you: no lists, no lurking.
What you will get
A fit snapshot: where your channel stands, and whether we are a match.
Two to three opportunities: specific, prioritised, yours to keep.
A recommended next step, even if that step is not us.
The audit is free and commits you to nothing: nobody follows up with a call you did not ask for.
The casework shows all of this done.
The case studies are the same work with the numbers, the sources and the dates attached.
See the casework →