UGC Creator Casting: What Actually Makes Someone Look Like Your Customer
Follower count predicts nothing about whether a creator's face makes your actual customer stop scrolling. Here's what does.
The most common creative brief we see when we take over an account is "find creators with good engagement." That's a real metric, but it answers the wrong question. Engagement tells you a creator's existing audience likes them. It tells you nothing about whether a stranger scrolling a cold feed will believe this specific person as a user of your specific product.
Why follower count is the wrong metric
A creator with 200,000 followers and a polished aesthetic often performs worse in cold prospecting than a creator with 4,000 followers who simply looks like the person buying your product. The cold audience isn't following the creator, they're evaluating whether this specific face and voice belongs to someone like them. Production polish can actively work against this: an ad that looks too produced reads as an ad, and the brain skips it before the pitch even lands.
The casting brief framework
Instead of briefing on follower tiers, we brief on four attributes tied directly to the buyer, not the creator's platform stats:
| Attribute | What it answers |
|---|---|
| Demographic proximity | Does this person plausibly belong to the customer's actual life stage, not an idealised aspirational version of it? |
| Problem credibility | Would this person plausibly have had the exact problem the product solves, in their own words? |
| Delivery register | Does their natural speaking style match how the target customer actually talks, not how a script sounds performed? |
| Existing trust signal | Is there any believable reason this person would know or recommend this category, without it needing to be stated? |
Matching creator to buyer psychology, not brand aesthetic
This is where casting briefs most often go wrong: optimising for who looks good next to the product instead of who the buyer would believe. On the skincare brand account, the winning angle for six months running was a creator reading her actual dermatologist bill next to the product, not a beauty-influencer aesthetic at all. The insight behind it: buyers weren't choosing skincare, they were leaving a dermatologist they couldn't afford. The creator's job was to embody that exact frustration, not to look like a skincare expert.
Rights and usage, sorted before the shoot
Usage rights are the part that derails scaling later if it's skipped early. Confirm paid usage rights, duration, and platform scope in writing before the shoot, not after a winning ad needs to keep running past an informal agreement's expiry. This is boring, and it's the difference between a winning creative you can scale for months and one you have to pull the week it started working.
What this looks like as a repeatable system
A pet care brand we work with runs casting as a standing process, not a one-off shoot: 25 to 35 new assets enter testing every month against a 160-asset owned library, and creator-shot footage consistently beats polished brand edits because it reads as advice from a pet parent, not a product announcement. That's the whole thesis: the creator's job is to be believable, not impressive.
Need creators cast to your actual buyer, not your mood board?
We script on buying psychology and cast for credibility, not polish, then monitor fatigue and iterate before performance drops.
Once creative is casted and scripted well, the next discipline is testing cadence, covered in how to spot creative fatigue before it shows up in ROAS.
