Supplement Marketing in AI Search: Our Benchmark Shows the Category Is Wide Open

Supplement marketing has a blind spot most brands haven't noticed yet. When a shopper asks ChatGPT, Gemini or Google's AI answers which magnesium to take for sleep, or what actually helps with menopause brain fog, the median supplement brand gets named in 5% of the answers. Across every other industry we measured the figure is 15%, which is low enough, but supplements sit at the very bottom of the table.

We know this because we ran the shopper questions ourselves. In October 2026 we put the questions real customers ask to five AI assistants on behalf of 135 online brands, read all 12,496 answers, and recorded who got named, in what position, how they were described and which websites the assistants cited. This post is what came out of it, with the supplement numbers pulled out on their own because that's where the gap, and the opportunity, is biggest.

If you're new to the idea of optimising for AI answers, our guide to answer engine optimisation for ecommerce covers the basics, and AEO vs SEO explains what changes for a Shopify store.

What is AI search visibility?

AI search visibility is the share of AI answers that name your brand when a shopper asks a question about what you sell. We measure it on non-branded questions only, the ones that don't mention you by name, because that's where a brand has to earn its place in the answer rather than being handed it.

Every brand in the benchmark was tested the same way:

Do AI assistants recommend online brands?

Mostly not. AI assistants know these brands, they just don't recommend them: ask an assistant about a brand by name and it recognises it in 100% of answers (median), but ask it what to buy in that brand's category and the same brand turns up in 15%.

A few more numbers that held across the sample:

• 63% of non-branded shopper questions never named the brand on any of the five assistants.
• When brands are named, AI speaks well of them. Of 3,344 mentions we classified, 13 were negative, which is 0.4%. The problem is presence, not perception.
• About 30% of answers now include shopping product cards, and the median brand appears in 3% of them.
• Shops and marketplaces make up 45% of the sites AI cites. Brands' own websites make up 5%.

Which industries get named most in AI search?

Sports and outdoor, food and drink, gifts and electronics brands get named most, in roughly a third of AI answers. Supplements, beauty, fashion and footwear get named least, and supplements is the lowest of the 14 industries we measured. The pattern is telling: categories where buyers lean on enthusiast knowledge do well, and categories dominated by big retail chains and household names do badly.

The industries at the bottom share a profile: big retailers carry the range, household-name brands fill the answers, and the questions shoppers ask are broad ("best collagen supplement", "what helps with bloating") rather than tied to a specialist product only a handful of brands make.

Why does supplement marketing have the biggest gap in AI search?

Supplement brands are named in 5% of AI answers because retailers and medical sources fill the answers instead, and the gap appears at every stage of the buying journey. When a shopper hasn't picked a brand yet, the median supplement brand is named in 1.8% of answers, and 81% of supplement questions never name the brand at all. Nearly two-thirds of the supplement brands we tested were under 10% visibility.

Retailers, not brands, own the answers

In many supplement categories the name AI reaches for first isn't a brand, it's a retailer. Boots, Holland & Barrett and Amazon lead more of the supplement competitor sets we tracked than any single brand does, and retailers and marketplaces take a median 28% share of voice. AI answers a supplement question by pointing to the shop that stocks everything, because no single brand has given it a better reason.

AI answers supplement questions from evidence

Supplement questions are health questions, and AI treats them that way. Medical and institutional sources make up 19% of the citations in supplement answers, more than double the 9% average across all industries, and PubMed and its PMC archive are the most cited sites in the category. Brands' own sites make up only 4%.

That tells you what an assistant is looking for before it names a supplement brand: a clear claim, backed by research it can trust, on a page it can read.

Shopping cards show up, supplement brands don't

Product cards appeared in 21% of supplement answers, but the median supplement brand appeared in none of them. Those slots went to retailers and marketplace listings. As AI shopping grows, and ChatGPT already shows product cards for shopping questions, that's revenue going to whoever has the cleanest product data, not the best product.

Why does being first in your category matter in AI search?

Because the most-named brand gets about four times the share of the typical brand. In each brand's competitor set, the most-named brand takes a median 23% share of voice, while the typical brand in the set takes 6%, and in AI answers that compounds: assistants tend to repeat the brands credible sources already agree on.

In supplements, the leader of a set is often a retailer holding around a quarter of the voice, and no brand has a dominant share. That's a fragmented category with no clear brand answer, which is exactly where one brand can become the default recommendation. The first supplement brand to give AI a clear, evidence-backed answer for "best magnesium for sleep" or "what helps with perimenopause" gets the slot the retailers are holding by default.

It's the same dynamic as early Google SEO, with one difference: AI gives a single answer with three or four names in it, not ten blue links. Being fifth is close to being invisible.

How can supplement brands get recommended by ChatGPT and other AI assistants?

Supplement brands get recommended by measuring the questions shoppers ask, publishing evidence-led pages that answer them, and getting onto the retailer and editorial pages AI already cites. These are the six moves we saw separate the brands that get named from the ones that don't, in the order we'd start.

1. Measure the questions, not the keywords. Start with the questions your customers actually ask an assistant, by stage: discovery ("what helps with…"), recommendation ("best X for Y") and branded. You need to see which assistants name you, which name a retailer instead, and which sites they cite. That's what we built StoreHero AI Search to do, running your questions through ChatGPT, Google AI Mode, AI Overviews, Gemini and Copilot every week.
2. Publish evidence-led ingredient pages. One page per ingredient and per need, answering the question in the first two sentences, then citing the studies behind it. AI cites research for supplement questions, so give it research it can attribute to you. Keep every claim inside what you can substantiate; the FTC's health products guidance is the standard to write to.
3. Answer the discovery questions. Supplement brands are named in 1.8% of discovery answers, the lowest of any stage in any industry. Symptom-led guides ("what helps with menopause brain fog", "magnesium for leg cramps at night") are where shoppers start, and where almost no brand is showing up.
4. Get onto the pages AI already cites. Your retailer listings, editorial roundups, comparison sites and community threads on Reddit and YouTube feed the answers. A complete, well-reviewed listing on the retailer AI names is often faster than waiting for your own site to be cited.
5. Fix your product data for shopping cards. Accurate titles, ingredients, dosages, prices and reviews in your product feed are what product cards are built from. Our guide to how ChatGPT picks products covers the feeds and crawlers each assistant uses.
6. Prioritise by margin, not by search volume alone. Push the products that make you money. Work out the contribution margin of each product first, and point your AI search work at the ones where an extra customer is worth the most.

What's next for the AI search benchmark?

This benchmark covers visibility, share of voice, sentiment and citations from one week of checks. It doesn't yet put a number on demand: how many people search for each supplement category every month. We'll add category search volumes in November, which will turn the visibility gap into a figure for the customers each point of visibility is worth.

Some caveats are worth keeping in mind. AI answers change week to week, so treat differences of a few points as noise. Competitor sets were chosen per brand, so share of voice describes the brands each one competes with directly, not the whole market. And the benchmark includes brands from several countries, so the industry figures blend markets together.

For more on where AI shopping is heading, see our breakdown of AI shopping agents like Meta Muse and whether ChatGPT ads are worth testing yet.

Key takeaways

• AI recognises brands when asked by name (100% median) but names them in only 15% of shopper questions that don't mention them.
• Supplement brands have the lowest AI visibility of 14 industries: 5% overall and 1.8% before a shopper has picked a brand.
• Retailers like Boots, Holland & Barrett and Amazon lead more supplement competitor sets than any single brand.
• Medical and institutional sources make up 19% of supplement citations, double the all-industry average, so evidence-led pages win.
• The most-named brand in a category takes about 4x the share of voice of a typical competitor, and no supplement brand has claimed that spot yet.
• Only 0.4% of AI brand mentions were negative: the problem is being named at all, not how brands are described.

Frequently asked questions

Sources

• StoreHero AI Search Benchmark, October 2026 (12,496 answers from ChatGPT, Google AI Mode, AI Overviews, Gemini and Copilot)
• OpenAI Help Center: Shopping with ChatGPT Search
• FTC: Health Products Compliance Guidance