Inside Sellable's AI Actor Library: 360+ Situation-Based UGC Scripts

The problem with one generic prompt

Most AI UGC and avatar-video tools work the same way: pick an actor, type a rough idea of what you want them to say, and generate. It works, but the output tends to converge on the same flat delivery regardless of what the product actually is — a person standing in a plain room, talking at the camera.

That's fine for a first test. It stops being fine once you're generating dozens of ads and every single one has the same generic setting and the same generic "hold up the product and talk about it" direction.

How the actor library is organized

Sellable's AI Actor library is built around situations, not just actors. Open the picker and you get a proper filterable browse experience — gender, age bracket, and 36 real scenario tags (car, gym, kitchen, beach, coffee shop, GRWM, unboxing, office, pool, and more), the same category of filtering a stock-footage library uses, rather than a flat scroll of thumbnails.

Pick an actor, pick a situation, and the library hands you one of that situation's pre-written scripts — 10 per scenario, 360+ in total — instead of a blank prompt field.

What each script actually contains

Every script in the library follows the same consistent structure, written specifically for AI video generation rather than adapted from a general copywriting template:

• Camera — shot type, framing, and handheld quality (iPhone-style selfie framing, natural micro-shake, no gimbal)
• Environment — the specific setting, described in enough detail to render accurately (a parked car with soft daylight through the windshield, not just "a car")
• Performance — how the actor should deliver the line: natural pacing, small pauses, genuine reactions, not exaggerated influencer energy
• Product — explicit direction for how the product itself should look and sit in frame (more on this below)
• Visual style — raw, unpolished phone-video aesthetic rather than a commercial look
• Script — the actual spoken lines, written as natural, slightly imperfect conversation rather than an ad read

That's a materially different starting point than a single open text field, and it's why the output doesn't default to the same generic delivery across every generation.

Product-placement direction, not an afterthought

The Sellable Platform

AI Product Photography for DTC Brands

Sellable turns a single product photo into studio-quality images, cinematic video, and on-brand campaigns — in under 60 seconds.

Get started free →

My Brand · Studio Generate

The detail most other AI UGC tools skip entirely: how the product itself should look on camera. Sellable's scripts include explicit direction that the product must match the real reference photo exactly — same packaging, color, and branding — and that it should sit naturally in the scene (held, set down on a counter, propped against something) rather than centered and floating like a studio ad insert.

The goal is a product that reads as something the creator actually owns and is showing off, not a commercial cutout pasted into a UGC-style shot. It's a small detail in the prompt, and it's the difference between video that feels like an ad and video that feels like a recommendation.

For teams evaluating scripts before committing credits to a full video generation, the library also supports generating a still-frame preview of a script's opening scene first — a quick way to sanity-check the setting and mood before running the full video.

*"The first time I used a competitor's tool, every video came out the same — person, plain background, holding the product up. Sellable's library is the first one that actually varies the setting."* — Nadia F., DTC wellness brand

Browse the AI Actor library and generate your first video →

FAQ