SolutionsCreative testing
Hooks, creators and formats, tested one axis at a time
hypjam writes the hooks, casts the creators, cuts the formats and lays them out so your ad account can test one thing at a time. Then it reads the grid and briefs the next round. Your team keeps the spend.
Ships to
- + your own site
A test needs variables you can name. The brief names them.
On hypjam hooks are tested against hooks, creators against creators and formats against formats, one axis at a time, so the winner means something. Four services make that a grid rather than a pile.
8 hooks
Per script, tested against each other
2 to 5 creators
Cast per brief, by package
1 debrief
A month, read back into the next brief
The services behind it
The services behind the grid.
Each one is a service on its own. Together they cover the variables, the people, the cuts and the reading.
- 01Hook writing
Eight hooks per script, deliberately different, so the test has a range.
Service page - 02Casting & roster
Two to five creators per brief, cast to differ on one thing at a time.
Service page - 03Editing & captions
Formats cut per placement so the account is comparing like with like.
Service page - 04Reporting
Reads the grid with you and writes the next brief from it.
Service page
Eight hooks, deliberately different, so the test has a range.
A test of three hooks that all say the same thing tells you nothing. The eight are written across kinds: a question, a confession, a comparison, a plain claim, a 'wait for it'. The shoot covers the ones you pick with us; the grid is built from them.
One axis per round
Round one tests hooks with the creator and format held steady. Round two holds the winning hook and tests creators. The brief says which axis is moving.
Written to be swapped
Each hook is written so it can sit on the same body of footage. A hook swap is the cheapest new ad an account can get.
Kept for the next round
Hooks that did not make the shoot stay in the brief for round two.
Two to five creators per brief, cast to differ on one thing.
Starter casts two creators per brief, growth four or five. They are cast to differ on the thing the round is testing: age, tone, setting, niche. You approve the shortlist, and each creator shoots the same hooks so the comparison is fair.
Cast on the axis
If the round tests tone, the creators differ in tone and match on everything else. If it tests setting, the kitchen and the car get the same script.
Same hooks, each creator
Every creator in the round shoots the same picked hooks, so a creator winning means the creator won, not the line.
Paid by us, once
Creators are paid out of your package fee on a set schedule. One contract, one invoice.
Formats cut per placement, so the account compares like with like.
A hook that wins on TikTok and loses on Reels may just have been cut wrong for Reels. Every variant is cut native for the placement it is testing in, captions burned in, so the format is a variable you control rather than noise.
Named variants
Each cut is named for its grid position: hook, creator, format, placement. The ad account's naming matches the brief, so the read is quick.
Native, not resized
9:16 for the placement, hook in the first frame, captions inside the safe area. Six placements, six cuts if the round needs them.
One revision round, included
Included on every package. Usually spent on captions, so that the caption is not the thing being tested by accident.
The grid is read with you, and the next brief is written from it.
On Growth and Scale, the month closes with a debrief: which hooks held attention, which creators carried the product, which formats the account should keep. It is not a deck for its own sake; it is the first page of the next brief.
Your numbers, our read
The performance data stays in your ad account. hypjam reads it with you and turns it into decisions: carry, retire, swap, scale.
Written into the brief
The next round's brief opens with the read: two more like hook 07, bring back the kitchen creator, drop the long format.
Honest about noise
Small tests are noisy. When a difference is not big enough to act on, the debrief says so rather than inventing a winner.
Built for the problems a testing budget actually has.
Not a pile of videos with a spreadsheet after. hypjam is built around the test that changed three things at once, the winner nobody can explain, and the debrief that never reached the next brief.
Three things changed at once
A new hook, a new creator and a new cut in the same ad means the winner cannot be explained. The brief moves one axis per round.
The winner nobody can explain
Named variants that match the ad account's naming make the read a five-minute job instead of an archaeology dig.
The debrief stayed in a deck
A read that does not change the next brief was a report, not a test. On Growth and Scale, it is the first page of round two.
Too small to call
Some rounds do not produce a clear winner. The debrief says so, and the next round is designed to find one.
How it runs
From the first variable to the next brief.
- 01
The axis is chosen
Hooks first, usually. The brief names what is moving this round and what is held steady.
- 02
Hooks and creators
Eight hooks written, the picked ones shot by every creator in the round, cast to differ on the axis.
- 03
Cut and named
Variants cut native per placement and named for their grid position, matching your ad account.
- 04
Read, then briefed
The month closes with a debrief on Growth and Scale, and round two opens with it.
Under the hood
One axis at a time, named for the account, read honestly.
The details a growth lead asks about before committing a testing budget.
Hooks
8 per script
Written across kinds so the round has a range. The picked ones are shot by every creator in the round.
Creators
2 to 5 per brief
Two on Starter, four or five on Growth, a dedicated roster on Scale. Cast to differ on the axis being tested.
Naming
Matches your account
Every variant is named for hook, creator, format and placement, in the convention your ad account already uses.
Debrief
Monthly
Included on Growth and Scale, written into the next brief. Starter includes the live tracker and our notes on what to run where.
Data
Stays with you
Performance data lives in your ad account. We read it with you; we do not hold it, and this site has no trackers.
Rights
Organic and paid
Every variant is licensed for organic and paid for the term, in writing with the creator before the shoot.
Rights and safety“We started hypjam because brands kept hiring creators one by one and nobody was directing the work. We treat UGC like a production: one brief, one team, a script before anyone films, an edit built for the feed.”
Avi Aggarwal
Founder, hypjam
FAQ
Questions growth leads ask first.
How many variants does a round produce?
It depends on the package and the axis. A Growth brief with three picked hooks, four creators and two placements is twenty-four named variants from one shoot; the brief sets the number before anyone films.
Do you run the ads?
No. The spend and the account stay with your team. We deliver named variants cut per placement and read the results with you.
What if there is no clear winner?
Then the debrief says so. Small tests are noisy, and inventing a winner would make the next round worse. Round two is designed to separate the close ones.
Can we test with our own creators?
Yes. They shoot the same picked hooks as the roster creators in the round, on the same contract, so the comparison is fair.
How is a testing round priced?
By package: starter from £1,800 a month with six videos and two creators, growth from £3,600 with twelve videos and four or five creators. The final number is scoped once we know the grid.
Tell us what you want to learn.
Book a thirty-minute call. Bring the ad account and the question; we will come back with the first round's axis, hooks and creators.
