Why Famous Brands Have Low AI Visibility

I ran 31 direct-to-consumer brands through our AI-visibility audit - ChatGPT, Claude, Gemini and Perplexity, across three categories. Going in, I assumed the brands with the loudest marketing would dominate. The opposite happened. The most famous names in the set are sitting on the floor.

Liquid Death scores 18. AG1 scores 20. Fenty Beauty scores 21. These are some of the most marketed DTC brands on earth, and the AI engines barely recommend them. Meanwhile CeraVe - no celebrity founder, no viral stunts - outscores Fenty two to one.

From what I have seen in our audit data, AI recommendation does not track fame, ad spend or cultural noise. It tracks something else entirely, and most brands are spending against the wrong thing. This post is the why, the proof, and the ten-minute fix.

AI visibility is a short list - and most brands miss it

First, the shape of it. Across 31 brands, 20 scored below 40 out of 100 - Limited or Low visibility. Only two cleared 80. AI is not handing out recommendations broadly. It keeps a short list per category and repeats it.

How short? In Sports & Fitness, the top two brands - Lululemon and Nike - both score a perfect 100. Every other sports brand in the set averages 36. That is a 64-point cliff between the two incumbents and everyone else. Same pattern, softer, in the other categories:

Notice who the top two are not. In Beauty, the engines' favorites are ILIA and CeraVe - not Fenty, not Rare Beauty, not Glossier, all far more famous. In Food, it is Four Sigmatic and Poppi, not AG1 or Liquid Death. The list AI keeps is not the list the culture keeps.

The famous brands with almost no AI visibility

Here are the brands that should win on name recognition alone, and what they actually scored:

Liquid Death has built one of the most recognized brand identities of the decade. AG1's ads are inescapable. Fenty and Rare Beauty are celebrity-founded juggernauts. Every one of them is, in AI terms, nearly invisible. Whatever they are buying, the engines are not reading it.

Why CeraVe beats Fenty in AI recommendations

Here is the mechanism, and it is the whole game. AI engines do not learn brands from billboards or a viral campaign. They learn them from text - the millions of articles, reviews, forum threads and "best X for Y" roundups they trained and retrieved on.

CeraVe wins because it is the answer to a written question. Search "best moisturizer for dry skin" and a decade of dermatologist roundups, review sites and Reddit threads name CeraVe. That density is exactly what the engine ingests. Fenty's coverage is enormous - but it is celebrity coverage, launch coverage, culture coverage. Almost none of it is the plain sentence "the best foundation is Fenty," written by a third party, ten thousand times over.

Brand mentions, not fame, drive AI visibility

This is not just a quirk of our 31 brands. Ahrefs studied 75,000 brands and found web mentions correlate with AI visibility at 0.664, while backlinks manage only 0.218 - being talked about beats being linked roughly three to one. In that same data 82% of AI citations come from earned media, and brands in the bottom half of web mentions are essentially invisible to the engines no matter how good their classic SEO is. Fame that never converts into earned, third-party text simply does not register.

This is the Mention Density Model in one example: recognition is a function of how densely you are referenced as the answer, not how famous you are. Princeton's Generative Engine Optimization research (Aggarwal et al., KDD 2024) found the same lever from the other side - adding citations, quotations and statistics to the source text an engine reads lifts that source's visibility by up to 40%. It is why a dermatologist-driven brand beats a celebrity one, and why a water company everyone can name cannot crack 20.

Why some product categories cap lower on AI visibility

The same mention-density mechanism explains the category ceilings. Mean visibility ran Sports & Fitness 47, Cosmetics & Beauty 35, Food & Nutrition 25. Functional food and supplements is the hardest room because the editorial answer space is thin and cautious - fewer trusted "best supplement" roundups exist, and the ones that do hedge on health claims. Less written density, lower ceiling.

How to get your brand recommended by AI engines

If fame does not buy AI recommendation, then marketing budget pointed at awareness will not move your score. Here is where it actually moves, and none of it is glamorous:

1. Get written up as the answer, by other people. Third-party "best X for Y" roundups, review sites, expert lists. With 82% of AI citations coming from earned media, one editor calling you the best is worth more than a million impressions you bought.
2. Feed reviews, not just ads. Dense, specific customer reviews and ratings are corpus the engines read. CeraVe's review depth is a moat; a celebrity face is not.
3. Win category-aided queries, not fame. You will not out-famous Nike. You can be the named answer to "best [specific category] for [specific use]" where the incumbents stay vague. Pick the room you can own.
4. Build the entity, so the engine is sure who you are. Consistent facts, Organization schema, Wikidata, sameAs. Fame is fuzzy to a model; a clean entity graph is not.
5. Measure recommendation, not awareness. Your brand-lift survey says you are famous. The AI audit says whether you are recommended. Track the metric that ends in a sale.

The brands winning AI search are not the ones spending the most to be known. They are the ones the web already treats as the answer. That is buildable - but you build it off your own domain, in everyone else's text, one earned mention at a time. Fame is a shortcut that does not exist here.

Check where your brand stands in AI search

Here is the move, and it takes ten minutes. Run GEOlikeaPro's Visibility checker across all four engines, then run the competitor you assume you are beating right next to it. Read the two scores side by side.

Most teams find the gap is not where the marketing dashboard says it is. The brand-lift survey says you are famous; the audit says whether you are recommended - the two different lists from the top of this post. Only one of them ends in a sale.

If the score comes back low, it is not a final grade, it is a worklist: the roundups you are missing, the reviews you are thin on, the entity facts the engine is unsure about. Fix those and the number moves. (For why a single engine is never enough to trust, see why the four engines barely agree.)

Sources. First-party GEOlikeaPro audit of 31 DTC brands across ChatGPT, Claude, Gemini and Perplexity. External: Ahrefs, "An Analysis of AI Overview Brand Visibility Factors" (75,000 brands, 2025); Aggarwal et al., "GEO: Generative Engine Optimization," KDD 2024; Mention Density Model.

Latest brands checked with our AI tools

The newest public reports from our AI visibility audit, hosting checker, and AI readiness checker - run yours to add your brand.