Guide
The two ways to brief a UGC creator badly: micromanage every word until it sounds like a script, or hand over the product and hope. The job is to control structure and free the voice.
The reason UGC ads work is that they feel like a real person talking, not a brand broadcasting. The moment you script every line, you lose the thing you paid for. But "just be yourself" produces a rambling video with no hook and a buried CTA. The fix is to brief the skeleton — the order of beats and the job each one does — and let the creator fill it in their own words. Control the architecture; release the delivery.
1. The hook. Write the literal first line or opening visual — this is non-negotiable and the part most briefs get wrong. 2. The structure. A simple, proven shape: relatable problem → product as the turn → quick demo or proof → clear CTA. 3. The message and claims. The one thing that must land, plus exactly which claims are allowed (keep it honest and compliant). 4. Format specs. Vertical, sound-on, captions, and length — for short-form, see how long an ad should be. 5. Do's and don'ts. A short list: shoot in good light, no logos in the first frame, don't open with the brand name.
The first three seconds decide whether the other twenty-seven get watched, so the hook deserves more of your brief than anything else. Don't write "grab attention" — give the creator two or three specific openings to choose from or reshoot: a bold claim ("I returned three of these before I found one that works"), a relatable problem stated out loud, or a pattern interrupt (an odd action before any words). Better yet, ask for the same video with three different openings — same body, different first three seconds. Reshooting only the hook is cheap for the creator and gives you a menu to screen. The deeper craft of openings is in how to write a TikTok hook.
A brief is a bet on what will work; the delivered video is the result, and the two don't always match. Before you put paid budget behind a creator's cut — or pick which of their hook variations to scale — screen them. PreTestAds' AdCortex™ engagement model predicts attention and engagement from the footage itself and scores each version on Hook Strength, Attention Drop, Peak Moment, and Purchase Signal, ranked against top-performing ads. That turns "which of these five do I like?" into "which of these five holds attention?" — and it tells you which creators to re-book. It slots straight into a small-budget testing framework: brief widely, screen the deliverables, spend on the winners. If you plan to run the creator's post through their own handle, the same screen applies to influencer whitelisting — test the cut before you put budget behind it. (Worth knowing the other side of the table, too: sharp creators price usage and whitelisting as separate line items — and the ones who show up with their own attention scores are usually the ones worth re-booking.)
Creators do their best work when the brief is short enough to hold in their head while filming. One page: the hook options, the four-beat structure, the message, the specs, and the do's and don'ts. Anything longer and the spontaneity that makes UGC convert gets edited out before the camera even turns on.
Upload your creators' deliverables and rank them before you spend — up to 8 free credits, no credit card.
Score My UGC AdsA strong UGC brief specifies the hook (the exact first line or visual), the rough structure (problem, demo, payoff, CTA), the key message and any claims you can legally make, format and length, and a few do's and don'ts. It controls the structure while leaving the creator's voice and delivery free.
Give the creator two or three hook options written as the literal first three seconds — a bold claim, a relatable problem, or a pattern interrupt — rather than a vague 'grab attention.' Specifying the opening is the single highest-leverage instruction in a UGC brief, because it decides whether anyone sees the rest.
Ask for three to five hook variations on the same core video — same body, different first three seconds. It's cheap for the creator to reshoot only the opening, and it gives you multiple openings to screen and test, which is where most of the performance difference lives.