Here’s a distinction most founders don’t realise exists until it costs them: a compliant formulation and a compliant marketing campaign are two completely different exercises, assessed by two completely different processes, and passing one tells you nothing about whether you’ve passed the other.
We see this constantly. A brand comes to us with a spotless CPSR, a properly notified product, a Responsible Person in place, labelling checked and double-checked — genuinely, textbook compliant on paper. Then their TikTok caption says “clinically proven to erase fine lines” and their Amazon bullet points promise it’ll “cure” hormonal acne, and none of that spotless paperwork protects them from what happens next.
Because the product being safe to sell and the claims you make about it being legal to say are governed separately. One sits under cosmetics regulation. The other sits under advertising law. You can nail the first and still get burned by the second.
The claim doesn’t live on the packaging anymore
This is the part that catches people out. A decade ago, the risk zone for a dodgy claim was fairly contained, your packaging, maybe a print ad. Job done, tick the box, move on.
Now your claims are scattered across a dozen surfaces you don’t fully control:
- Your website copy, the most “official” version, and usually the most carefully checked
- Instagram captions and Stories, written fast, often by whoever’s running the account that week
- TikTok videos, where a throwaway line in a voiceover carries the same legal weight as something printed on a box
- Amazon listings- bullet points, A+ content, and the product title itself, which brands routinely stuff with claim-heavy keywords to win the algorithm
- Influencer and UGC content- and here’s the bit people really don’t want to hear: if an influencer you’ve paid or gifted product to makes an exaggerated claim about your product, you can still be held responsible for it, because you commissioned the content
Every one of those is a live claim. Every one of them can be reported, screenshotted, and investigated. A regulator doesn’t care that the offending line came from a 15-second TikTok clip rather than the box; the claim is the claim, wherever it lives.
Who’s actually checking, and what they’re checking against
In the UK, that’s the Advertising Standards Authority (ASA), working from the CAP Code — the Committee of Advertising Practice’s rulebook for non-broadcast advertising, which covers websites, social media, and online marketplaces, not just billboards and TV. The two rules that come up again and again are 3.1 (don’t mislead) and 3.7 (substantiate what you claim, before you claim it, not after someone complains). The ASA doesn’t need a formal complaint to open an investigation, and platforms increasingly cooperate with takedown requests once a ruling lands.
And this isn’t a small-brand problem you can afford to feel above.
Even L’Oréal gets caught out
If you think a brand with an entire in-house legal and regulatory team is somehow immune, look at the record. L’Oréal, the biggest cosmetics company in the world has been repeatedly pulled up by the ASA, and not for obscure technicalities.
Take Garnier’s Vitamin Cg Serum. An advert for it aired in late October 2025 carrying the claim “clinically proven to reduce hyperpigmentation in 2 weeks.” The ASA investigated whether that claim could actually be backed up and banned the ad. L’Oréal’s own explanation was almost worse than the ruling itself: the small-print qualifier attached to the claim referenced the wrong study entirely, mixing up a subjective 72%-agreement survey figure with the objective clinical data that should have been cited instead. Not a rogue influencer, not a junior social media hire. The brand’s own advert, its own paperwork error, and still a ban.
It’s not an isolated case either. The ASA has separately taken action against L’Oréal, alongside other brands, over LED face mask adverts that strayed into making medical-sounding claims, and the company has been pulled up more than once over digitally altered campaign imagery that misrepresented what a product could realistically achieve on skin.
The point isn’t to pick on L’Oréal. The point is: if a company with that much money, that much legal oversight, and that much at stake still gets tripped up on a mislabelled qualifier, what chance does a claim written in a rush by whoever’s managing your Instagram this week actually have, if nobody’s checking it against the same standard?
What this actually means for you
Your regulatory file gets you a product you’re allowed to sell. It says nothing about the language you’re allowed to sell it with. Those are two separate approvals, and most small brands only ever build a process for the first one.
If your CPSR is airtight but nobody’s checking the claims going out on TikTok, Instagram, Amazon, and your influencer briefs against the same evidence standard, you don’t have a compliant marketing operation, you have a compliant product and an exposed one everywhere else.
