UGC style ads without a shoot

There is no button that turns a product into a converting advert, here or anywhere else. There is a short set of decisions, each with a right answer more often than people assume: which format you are performing, how long, whose face and voice, and what proof makes the claim believable.

Across 179 finished talking head renders, the median clip is 19 seconds and under 3% run past two minutes. For an ad, write twenty seconds: one line of hook, two or three of substance, one of exit, and read it aloud before rendering.

Use 4:5. It takes more of a feed than 1:1, survives being cropped to 1:1, and is not letterboxed the way full 9:16 can be when a platform shows it in feed.

The order that works is format first, then the words written to it, then the render. See remake a viral video and realtime AI video.

A brand market scan has an optional leg that pulls competitor creative from the Meta Ad Library. It shows what rivals pay to run, their creative and their claims. It does not carry view counts, so it cannot tell you what works. For ranking, use organic signals; both are covered in competitor video analysis.

The hook and the cutting can be replicated. The proof cannot, and it decides whether the ad works: a number you actually have, a before and after you actually ran, a real message from a real customer. A generated face reading a real result is a different object from a generated face making a claim.

A UGC style ad borrows the look of creator content. It must never present a generated person as a real customer or a genuine testimonial. Use real customer words only with permission, and disclose AI where it matters. See ethical AI video generation.