Verified against Veo 3.1 · 2026-07-31
Generate a UGC-style testimonial ad clip that reads as unproduced
A Veo 3.1 prompt for a vertical, selfie-framed testimonial clip that deliberately requests the imperfections of real user-generated content, since a produced-looking testimonial is the fastest way an audience discounts it as an ad rather than a review.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Write a Veo 3.1 prompt for an 8-second vertical, selfie-style testimonial clip in the style of authentic user-generated content — the specific genre defined by its imperfections, not despite them, so those imperfections need to be requested explicitly rather than left to a default that will smooth them out. SPEAKER a woman in her early 30s, hair pulled back, wearing a plain t-shirt, holding the camera at arm's length in a selfie-style framing, speaking directly to camera. SETTING her own kitchen, dishes visible on the counter behind her, an ordinary, real, slightly cluttered space — not a staged or styled backdrop, since a testimonial filmed in an obviously produced setting undercuts its own credibility before a single word is spoken. DIALOGUE Priya says: "Okay so I was, like, genuinely skeptical about this — but I actually use it every single morning now." Write it as a real person would actually speak, not as ad copy — include a natural filler word or a slight self-correction if it fits, since a testimonial that sounds too polished is one of the fastest signals an audience uses to discount it as scripted. CAMERA Handheld, held by the speaker themselves, with the natural micro-shake of an actual selfie-style shot — not a smoothed, stabilized gimbal move. State explicitly that stabilization should look like a person's own hand, not a locked or gimbal-smoothed frame, since Veo's default handheld interpretation sometimes skews smoother than genuine phone-in-hand footage actually looks. LIGHTING morning light from a kitchen window to her left, ordinary available light — a window, an overhead room light, nothing set up as a proper key light. Explicitly rule out a visible softbox or ring-light glow unless the described setting genuinely includes one, since an obviously lit face is one of the clearest tells that a "testimonial" was actually shot as a production. AUDIO a slight kitchen echo, a faint hum from a running dishwasher in the background, with the speaker's voice carrying the natural acoustic signature of the named room — some echo in a bare room, a flatter sound in a soft-furnished one — plus whatever ordinary ambient sound the space would actually have, not a clean, isolated studio voice track. STYLE AND MOOD slightly overexposed near the window, ordinary phone-camera color, no grade applied. Genuinely unpolished — slightly uneven exposure, a phone-camera color response, not a color grade. WHAT TO AVOID Do not ask for any element that reads as intentionally produced — a music bed, a lower-third graphic, a smooth camera move, or dramatic lighting — inside this generation; if a produced element is needed for the final ad, add it in post over this deliberately raw base clip, not inside the same generation that is trying to look unproduced. OUTPUT The finished prompt, followed by one line naming the specific imperfection chosen (a filler word, a hand-shake, room echo) that is doing the most work to sell authenticity.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
The entire premise of UGC-style advertising is that its credibility comes from reading as unproduced, which means every default Veo applies toward a cleaner, more polished result is actively working against the goal here, and the prompt has to counteract that default explicitly rather than just omitting production cues and hoping the result lands raw by default. Writing the testimonial line as real speech — including a filler word or a slight self-correction — rather than as tightened ad copy targets the specific way audiences detect scripted content: real spoken language is disfluent in small, consistent ways, and a testimonial that is grammatically clean and rhythmically even is one of the fastest tells that it was written rather than spoken, regardless of how casual the setting looks. Explicitly requesting the natural micro-shake of a genuine selfie-style shot rather than a smoothed handheld look matters because Veo's own interpretation of "handheld" sometimes lands closer to a stabilized gimbal move than to actual phone-in-hand footage, and that smoother default is exactly the kind of visual polish that reads as produced the moment it is set next to real UGC content in the same feed. Naming the room's acoustic signature — echo in a bare kitchen versus a flatter sound in a soft-furnished room — and pairing it with a clean voice track is a mismatch that real UGC never has: authentic phone audio always carries the room it was recorded in, and a clean isolated voice over a visually lived-in space is a subtle audio-visual inconsistency that undermines the illusion even when neither element looks wrong in isolation. Finally, explicitly deferring any genuinely produced element — a music bed, a graphic, dramatic lighting — to a later post-production pass over this deliberately raw base clip, rather than asking Veo to generate both the raw authenticity and the polished ad elements in the same shot, respects the fact that those two things are in direct tension: a single generation asked to be simultaneously raw and produced tends to land in an uncanny middle ground that convincingly reads as neither.
Verified against
Veo 3.1 3.1 · 2026-07-31
Changelog
- 2026-07-31 — Initial publish, verified against Veo 3.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
