In short: To test UGC ads, change one thing at a time, decide in advance when you will stop an ad, and name every file so a result can be traced to what was in the video. The hook, the first two to three seconds, is the cheapest thing to vary: TikTok’s split-test tool lists video hooks as a variable and allows one variable per test. Insense’s September 2026 workflow guide says to brief creators for modular footage, such as several hooks and call-to-action endings, so one filming session produces many variants. Platform vendors publish most of the figures below; none is independent research.
What you can test
| Variable | Example |
|---|---|
| Hook | Same video, different first 3 seconds |
| Creator | Same brief, different person |
| Angle | Texture vs routine vs result |
| Call to action | Same body, different ending |
| Length | 15 vs 30 seconds |
TikTok’s help center lists the variables its split test can change: targeting, placement, creative, budget strategy, creative assets (images, videos, formats and video “hooks”, the opening two to three seconds), catalog, and bidding and optimization. Its rule is plain: “You can only select one variable for each split test.” The reason is simple. If a test changes the creator and the hook and the offer at once, a winner tells you nothing about why it won.
Brief for variants, not for one video
Insense argues that scaling UGC “is a workflow problem, not a sourcing problem”, and that a single creator engagement “should produce a library, not a video”. It recommends asking for hooks, demos, call-to-action endings, b-roll and raw files instead of one finished cut. Its worked example is hypothetical: three hooks and two endings from one modular brief give ten or more variations. Billo, another platform, likewise suggests asking creators to film two to three hook variations of the same core video. Vidovo gives a similar example, from one filming session: three opening hooks, one demonstration and two benefit-focused versions.
That fits a one-page brief. Add a line such as: “Film the same demonstration three times, each with a different first sentence.” See the beauty brief template and UGC video ideas and hooks.
Name files so results can be traced
Insense suggests delivering files in folders for raw footage, hooks and b-roll, with names by component such as hook_01.mp4 and cta_01.mp4, and tagging assets by date, creator, hook type and call to action. Vidovo’s list of what to record next to each video runs longer: creator, hook, format, length, call to action, platform, type of use, spend, click-through rate, hold rate, conversions, cost per acquisition and return on ad spend. It says to judge performance by creative angle, not just by creator name. Choose a naming rule before the first video arrives.
Decide how you will stop an ad
Insense advises setting kill criteria in advance, so that nobody argues about a number after the money is spent. It also gives rules of thumb, such as looking at early hook and thumb-stop metrics within 24 to 48 hours, which it presents as guidelines. Treat these as starting points. Guides differ on how long a test must run: the sources behind how to measure UGC performance suggest at least seven days for a simple comparison, while TikTok’s own advice for formal split tests, published in 2021, is a minimum of two weeks and perhaps up to six. TikTok also describes testing as an ongoing habit: “It’s not about solving the problem, it’s about getting a little better each time.” Check the current guidance for the platform you use before you set a rule.
How many videos to brief
Billo suggests testing volumes by monthly ad spend. These are its own baselines, not independent research:
| Ad spend per month | Billo’s testing volume |
|---|---|
| Under $50,000 | 10–24 videos a month |
| $50,000–$100,000 | 25–50 videos a month |
| Over $100,000 | 25–50 variations a week |
Billo gives roughly $198 as an average cost per UGC deliverable, from its own platform data, and works out $1,980 to $4,752 for 10 to 24 videos. Prices on other platforms differ; see what UGC creators charge and how many UGC videos you need. Small brands can test with far less: two or three hooks on the best video is a real test.
A testing checklist
- One variable per test, written down before the ads go live.
- A brief that asks for several hooks or endings from the same filming session.
- A naming rule that records creator, hook and call to action.
- A stop rule agreed in advance, and a minimum run time from the platform’s own guidance.
- The result noted next to the creative, so the next brief starts from what won.
- License and usage rights confirmed before spend; see UGC usage rights explained.
On Usicri
A Usicri campaign collects videos from different creators, which gives you different faces, voices and hooks to test. You set the payout per video and how many videos you want, review each one and approve the ones that fit; you pay only for approved videos, with no platform fee at the moment. An approved, paid video comes with a non-exclusive, worldwide license without time limit that covers advertising (section 5 of the Terms of Service), so you can use it in your own ads. Ads run through a creator’s own account are not included. Usicri does not run or measure your ads, so the tests happen in your own ad account. A campaign is not guaranteed to receive submissions. See how it works for brands.
To pick which videos to test, see how to choose between UGC submissions, and to get more out of each one, read how to reuse one UGC video across channels.