If you type “best retinol for sensitive skin” into ChatGPT then you’ll get a shortlist of five or six brands by name. No sponsored row, or shopping carousel, and no way to bid your way onto it. Beauty has always run on somebody else’s recommendation: the friend who won’t stop talking about a particular serum, the beauty editor whose verdict you really trust, the Reddit thread you read doom scrolling at 11pm, or the pharmacist who pointed you towards the £6 version.
That AI recommendations are now doing the same job at this, but at scale. This means the job for beauty marketers has changed scope- how to get your brand or product to rank for a particular keyword or key phrase now matters less than being recommended in the answers.
Beauty shoppers are always asking someone else
The beauty category has never been a simple solo or impulse decision. People ask before they buy, and they ask because the stakes feel more personal with greater consequences – for example, will your skin breakout? A foundation shade or an acid that might wreck your barrier, a £48 serum you’ll be living with for three months but has reacted badly to your skin.
These discovery and purchasing habits are now colliding with a big change in how search works. Pew Research Centre found in July 2025 that when Google showed an AI summary, users clicked through to a traditional result only about 8% of the time, down from 15% when no summary appeared. The beauty industry feels this faster than most sectors because so many beauty queries are advice-shaped rather than navigational. Like, does this work on 4c hair? What’s serums are safe to use in pregnancy? Is the e.l.f. contour dupe as good as the original? An AI answer handles all of these queries in one go, and the researcher never clicks through, they never actually reach your product page.
The money involved makes it worth paying attention to. The British Beauty Council’s Value of Beauty 2026 report put the industry’s contribution to the UK economy at £28.3 billion in 2025. A meaningful slice of that gets decided before anyone lands on a website at all…
What AI reads when it builds a beauty shortlist
Editorial coverage is the raw material
Muck Rack and Generative Pulse’s 2025 “What Is AI Reading?” analysis found that more than 95% of links cited in AI answers were unpaid, and around 85% of those were earned media. Ahrefs, looking at 75,000 brands in 2025, found that branded mentions across the web correlated with AI Overview visibility at 0.664, which is a strong relationship by any standard.
Read that through a beauty lens and it’s pretty obvious.
The brands appearing in AI answers are the brands appearing in the likes of Cosmopolitan UK’s best-of roundups, in Get The Gloss ingredient explainers, in Refinery29 reviews and Which? Reviews testing.
Your paid social spend does not feed that machine. Your owned blog barely does.
Earned coverage does, which is why digital PR has quietly become a visibility channel rather than a brand-awareness nice-to-have it once was.
Beauty editors can tell when a pitch was written by a robot
Time pressure is a huge reason why AI-generated press releases may seem tempting. But the trade-off is worse than most people assume. Medianet’s 2026 Australian Media Landscape Report surveyed over 800 journalists and found 78% said AI-generated pitches lowered their trust in PR content, while 48% reckoned they could almost always spot an AI-written pitch.
Bare in mind, the same report found that 86% named press releases as their most-used story source, so the channel still works but it needs to be human-led. A features editor at a national paper who has had four generic “revolutionary skincare breakthrough” pitches from you is not opening the fifth, and no amount of volume will fix that, the interesting, unique angle will.
The angle is the thing AI can’t invent
Language models produce statistically likely phrasing, that’s why every AI-written beauty release reaches for the same vocabulary. Like revolutionary, game-changing, transformative, clinically proven or glow type phrases. Editors have read those words nine hundred times this month… and with the latest developments on watermarks in text, it’s only getting easier to prove this.
What earns coverage is material that only you have. Sales data showing a 300% spike in searches for barrier repair products after a particular TikTok trend, or your formulation chemist explaining why a trending ingredient doesn’t do what the internet says it does. Independent consumer research on how women over 45 actually shop, or a founder story with a real reason for existing behind it – these stories stand out.
Strong beauty digital PR is built on assets like these because they give a journalist something to write and therefore gives an AI something to cite. Ask yourself “what does your brand know or do that nobody elses does” then build the story out from there (rather than starting with the product).
A 20-minute test you can run before your next meeting
Here’s something to try – open ChatGPT, Gemini and Google’s AI Overviews and ask each one the questions your customers may ask. “Best affordable vitamin C serum UK” “what should I use for postpartum hair loss” “cruelty-free foundation for oily skin”
Then check three things:
- Is your brand named at all?
- Which publications are being cited in the answers?
- Are your closest competitors there, and what coverage put them there?
That citation list is a media target list someone else has already built for you. It tells you exactly which titles carry weight in your category, which is a far better starting point than a bought media database.
Where this leaves your PR plan
AI belongs in your PR workflo- use it to summarise research, tidy a media list, spot a trend early or proof a draft at 6pm. Hand it the parts of the job that were never the valuable bit.
You should be keeping the angle, the data interpretation and the journalist relationships firmly human, because those are the elements that turn a story into coverage and coverage into citations.
The beauty brands winning visibility right now are not the ones publishing most often, they’re the ones giving editors something genuinely worth writing about.
We recently conducted our own independent AI in beauty research which is available here. We did a deep dive analysing which brands are being recommended, where AI platforms are actually sourcing their answers, for example Who What Wear, and what the data reveals about the gap between traditional digital authority and real AI visibility. We hope this white paper will help you to get your beauty brand winning in AI Search.
If your brand isn’t showing up when an AI is asked what works in your category, that’s a fixable problem, and it starts with earning coverage that deserves to be there.
Read the white paper above, and as a beauty digital PR agency, Foundation can help you build campaigns and stories that beauty editors actually want to read and AI models will cite.
Take a look at our digital PR services and let’s talk about getting your brand into the answers.
