Use Seedance, Veo 3.1 & ChatGPT to Make UGC That Sells on Meta and Google Ads

I wrote the first version of this post in mid 2025 and most of it aged badly. The tools it was built on are gone or outclassed, and the current generation of video models got better at the exact things ad creative needs: product consistency, native audio, and cost per clip low enough to actually test at volume.

This is the fully updated version. Same idea, current tools. I run paid social for a DTC brand I run ops for, and this is the workflow we actually use to generate UGC-style creative without filming anything.

Why UGC still works in paid ads

Nothing here changed. UGC feels native in the feed, carries social proof, and outperforms polished studio creative on mobile-first placements. What changed is that you no longer need creators to make it, at least not for the testing phase. The bottleneck used to be waiting on creator submissions. Now the bottleneck is how fast you can write prompts.

The 2026 model landscape, quickly

You don’t need one model. You need to know which one to reach for per job.

Seedance 2.0 (ByteDance, released February 2026) is the workhorse for commercial creative. It sits at the top of the Artificial Analysis leaderboard, handles reference images so your product looks like your product across clips, does multi-shot continuity, and generates synchronized audio in the same pass. It’s also cheap: roughly $47 for a hundred 10-second videos. That’s the cost structure that makes real creative testing possible.

Veo 3.1 (Google) is the realism and prompt-adherence pick. It rated highest for following prompts accurately in MovieGenBench evaluations, generates 48kHz speech natively, and its Ingredients to Video feature locks character and product identity from reference images. Fast tier is $0.09 per second, Standard is $0.18. When a clip needs to feel indistinguishable from a phone recording, this is the one.

Kling 3.0 (Kuaishou) is the volume play. Around $0.095 per second on Pro, with lip sync in five languages. If you’re running Spanish-language creative alongside English, which we do, the multilingual lip sync matters more than any benchmark score.

Wan 2.7 (Alibaba) is the open source option if you want to self-host and drive cost to near zero at scale. More setup, more control.

And if you’re wondering about Sora: it’s dead. OpenAI shut the app down in April 2026 and the API goes dark on September 24, 2026. Don’t build anything on it.

1. UGC copy prompts

Copy still comes first, because the video models take direction from the script. These prompts work in ChatGPT, Claude, or whatever you use.

A. Testimonial-style caption

“Write a casual customer review of [PRODUCT] from a 28-year-old woman who bought it to solve [PROBLEM]. She was skeptical at first but now loves it. Make it sound like a real Facebook comment.”

Use for Meta primary text, Google Ads descriptions, overlay text. But read the warning section below before you run testimonial-style creative.

B. Unboxing first impressions

“Generate a natural, first-use reaction to [PRODUCT], with comments about packaging, smell, feel, or results. Make it feel unscripted.”

C. Problem-solution in 60 words

“Write a short UGC-style Facebook ad caption about how [PRODUCT] helped solve [PAIN POINT]. Include hook, struggle, solution, and result in under 60 words.”

D. Carousel voice variety

“Write 5 short UGC-style blurbs (1 to 2 sentences) from different types of users of [PRODUCT], each with a different personality: skeptic, enthusiast, quiet observer, and so on.”

2. UGC-style images

The image work now runs through Google’s Nano Banana Pro (the Gemini image model) or GPT Image. Nano Banana Pro is notably better at keeping a product render consistent across variants, which is the whole game for ad images.

The reverse-prompting workflow survived the model swap intact:

1. Upload a UGC photo that already performed well for you into ChatGPT or Gemini.
2. Ask:

“Give me a detailed prompt to generate an image similar to this. Same pose, lighting, background, and mood. I want it to look like UGC.”

3. You’ll get something like:

“Generate an image of a young woman smiling in her bathroom, holding a dropper bottle of serum. Handheld angle, soft morning light, natural skin texture, no makeup, casual loungewear.”

4. Run that prompt with your actual product image attached as a reference so the model doesn’t invent a fake bottle.
5. Generate variants: different people, AM vs PM lighting, slight pose changes, different product formats.

That gives you an image bank for carousel and static testing that looks real and on-brand. Here are two examples, and the current models beat these easily:

3. Full UGC-style videos with Seedance 2.0 and Veo 3.1

This is where the 2026 stack actually earns its keep. The old workflow was: generate silent video, record or synthesize a voiceover, sync it in an editor. The new models generate the person, the dialogue, and the ambient sound in one pass.

My default: Seedance 2.0 for product-in-hand clips where consistency matters, Veo 3.1 when the clip needs maximum “someone filmed this on their phone” realism.

Step by step

1. Write the video prompt in ChatGPT or Claude:

“Write a prompt for an AI video model to generate a 10-second UGC-style video for [PRODUCT]. Make it look like someone recording themselves on a phone camera, reacting casually after trying it. Include location, camera angle, lighting, props, and the exact line of dialogue they speak.”

Example output:

“A 30-year-old woman stands at her bathroom mirror holding a bottle of facial serum. She dabs it on her cheeks while speaking casually to the camera: ‘Okay wait, this actually feels amazing.’ Handheld phone camera angle, soft morning light, real cluttered counter in background, unfiltered skin texture, natural room audio.”

2. Run it in Seedance 2.0 or Veo 3.1 with your product photo attached as a reference image. This step is not optional. Without a reference, every clip shows a slightly different invented product and none of it is usable.
3. The dialogue and audio come out of the model already synced. No editor pass needed for the raw clip.
4. Use the output as Reels, Stories, YouTube in-feed, or Shorts creative.
5. Layer text overlays generated alongside the script:

“Write 3 on-screen text overlays for a UGC video about [PRODUCT] being shockingly effective after just one use.”

For Spanish variants, rerun the same prompt through Kling 3.0 with the translated dialogue line. The lip sync holds.

4. Assemble the funnel

Same structure as always, just faster to fill:

• Top of funnel: AI UGC video in Reel and Story format, Seedance or Veo output
• Middle of funnel: carousel of UGC-style images from Nano Banana Pro
• Bottom of funnel: retargeting statics with text overlays and native-sounding captions

The real unlock isn’t any single clip. It’s that at roughly fifty cents per 10-second Seedance clip, you can test 20 hooks against each other in an afternoon for the price of one creator brief.

One warning before you scale this

Don’t present AI people as real customers. An AI-generated woman saying “I bought this and it changed my skin” framed as a genuine review is a fake testimonial, and the FTC’s fake reviews rule covers exactly that. It’s also just a bad trade: the first commenter who clocks it as AI torches the ad’s social proof.

What works and stays clean: script the clips as demos, skits, founder-style pitches, or obviously stylized creative. The UGC aesthetic (handheld camera, natural light, casual delivery) is what drives performance, not the false claim that a specific real person bought your product. You get the performance without the liability.

Nikhil Sharma

I'm Nikhil Sharma. I write about Shopify, paid ads, email, and the systems I build for the DTC brands I work with.