What to put in a UGC ad brief, section by section, with the fields that decide whether the first cut is usable. Includes a copy-paste template and the numbers on how much a clear brief saves.
HighQualityUGC Team
Editorial
We run UGC ad tests daily and publish what holds up: real credit costs, real hook rates, no vendor fluff.
Most bad UGC ads were bad before anyone pressed record. The creator delivered exactly what was asked for. What was asked for was "a fun 30 second video about our product, feel free to be creative."
That sentence costs money in a way that never shows up on an invoice. It shows up two weeks later as a video you cannot run, a revision round you did not budget for, and a test slot that stayed empty while you waited.
This post is the brief itself. Eight sections, what goes in each, what a filled-in one looks like, and where teams keep getting it wrong.
A UGC ad brief is a short document that tells whoever makes the video exactly what has to be true about the finished file: who it speaks to, what it claims, what has to be on screen, and what the viewer is asked to do.
The reason it matters more than the shoot is arithmetic. A revision round on a creator video is a week. A brief edit is four minutes.
With a structured brief that number reportedly falls to 15 to 20%. Same creators, same product, same budget. The only variable is how much of the thinking happened before the camera came out.
There is a second reason, and it is the one media buyers care about. An ad you cannot read cannot be tested. If three videos differ in hook, pacing, actor, framing and closing line all at once, and one wins, you have learned nothing you can reuse. The brief is where you decide what stays fixed.
Not a persona document. One sentence that names a person specific enough to disqualify people.
Bad: "women 25 to 45 interested in wellness."
Good: "someone who already takes a magnesium supplement, has tried two brands, and stopped because the pills were too big to swallow."
The second version writes the script by itself. The first version is a targeting setting that wandered into a creative document.
The problem in the buyer's own words, not the category's words. Buyers do not say "suboptimal bioavailability." They say "I take it every day and I do not feel anything."
Pull these from real sources: your own review inbox, support tickets, the one-star reviews of your closest competitor, the relevant subreddit. A line lifted from a review outperforms a line written in a meeting, because it was already proven to be how a real person phrases the complaint.
The single thing the ad argues. One. If the brief lists three benefits, the video will list three benefits, and a video that lists three benefits argues for none of them.
Write it as a sentence a viewer could repeat to a friend. "It is a powder, so there is no pill to swallow" is a claim. "Premium quality ingredients" is not.
The first two seconds, written out word for word, plus what is on screen while it is said.
This is the field teams skip most often and the one that decides the ad. Roughly two thirds of the people who ever see the ad will see only this part. Give the hook its own line in the brief, and give it a visual instruction, not just dialogue: a face, a product in hand, a mess on a counter, a before state.
If you are testing hooks, this is the only field that changes between variants. Everything else in the brief stays byte for byte identical.
The video in four to six beats with seconds attached. Not a script. A skeleton.
| Seconds | Beat | What is on screen |
|---|---|---|
| 0 to 2 | Hook, spoken | Face, handheld, product not visible yet |
| 2 to 6 | The problem, stated once | The failed alternative in hand |
| 6 to 12 | The claim, demonstrated | Product used, one continuous action |
| 12 to 17 | The proof or the objection killed | Close crop on the detail that proves it |
| 17 to 20 | One ask | Product held, label readable |
Seconds matter because attention is spent, not given. A beat sheet without timings turns into a 45 second video where the product appears at second 22.
What has to physically exist in the file. Three to six lines, each one binary: it is there or it is not.
This is the section that prevents the most expensive kind of revision: the one where the video is good and still unusable.
One action, spoken plainly, in the last three seconds. Not two. A second ask splits the response.
Write the exact words. "Link is below" and "go check it out if that sounds like your problem" perform differently, and the difference is not something the creator should be guessing at.
The commercial half, and the half that causes the arguments.
When the video is generated rather than filmed, the brief does not shrink. It gets used differently: instead of a person interpreting it, it becomes the input the storyboard is built from.
The advantage is that the revision loop collapses. A hook change is a re-render, not a reshoot, so the beat sheet and shot list stay fixed while the one variable moves. The failure mode also changes shape. A vague brief given to a person produces a well-shot video about nothing. A vague brief given to a model produces something that looks generic in a way viewers immediately clock.
Either way, the fix is upstream. Decide the buyer, the problem and the claim before anything renders.
Most briefs are too long and light on the fields that matter. Delete these:
A brief that fits on one screen and answers all eight sections beats a six page deck that answers three.
Eight sections: the one buyer described specifically enough to disqualify people, the one problem in the buyer's own words, the one claim the ad argues, the hook written word for word with its visual, a beat sheet with seconds attached, a shot list of pass or fail requirements, one closing ask, and the rights, revisions and turnaround terms.
One screen. A brief that answers all eight sections in under 400 words outperforms a multi page deck, because the fields that get skipped in long briefs are the hook and the shot list, which are the two that decide whether the video is usable.
One to two rounds is the common package, and the brief should define what counts. A fix inside the agreed beats is a revision. A change of setting, actor or claim is a new scope. Writing that distinction into the brief is what stops the argument later.
Reported figures put first submission revision rates around 55% for vague briefs versus 15 to 20% for structured ones. The mechanism is simple: most revisions are not quality problems, they are specification problems, and a specification problem can only be fixed before the shoot.
No. The same eight sections apply. What changes is that the brief feeds a storyboard you approve before rendering, so hook variants become re-renders instead of reshoots, and the beat sheet stays fixed across every variant you test.
