"Best for sensitive skin": how AI labels the products it recommends

Published

By Carson Cobb. Data collected November 4 and 5, 2026.

When ChatGPT recommends products, it rarely just lists them. It labels them: “Best overall”, “Best value”, “Best for dry skin”, “Best for big dogs that pull”. That label is often the first thing a shopper reads, and it decides which pick fits them.

I went back to the answers from my Who ChatGPT recommends research, 48 buyer questions asked in ChatGPT, Gemini and Google AI Mode, and counted every label each AI gave.

The short version

  • ChatGPT labels almost everything. 44 of 48 answers had at least one “best for” style label, 149 labels in total.
  • Gemini and AI Mode label less. 31 and 29 answers.
  • Most labels name a specific need. 45% of ChatGPT’s labels and 51% of AI Mode’s described a shopper situation, not “best overall”.
  • “Best overall” is one spot per answer. It went to about a third of ChatGPT’s labels.
  • Value and budget labels are the next biggest group. 17% of ChatGPT’s labels.

How often each AI labels picks

ChatGPTGeminiGoogle AI Mode
Answers with at least one label44 of 4831 of 4829 of 48
Total labels1496676

What kind of labels

Label typeChatGPTGeminiGoogle AI Mode
Specific need (“best for dry skin”)67 (45%)26 (39%)39 (51%)
Best overall or top pick51 (34%)28 (42%)22 (29%)
Value or budget25 (17%)11 (17%)11 (14%)
Premium or splurge6 (4%)1 (2%)4 (5%)

Examples of specific need labels from the answers: best for natural fiber comfort, best for muscular legs, best for extreme cold, best for oily skin, best for hot sleepers, best for hard water, best for septic systems, best for trail running, best for gifting.

What this means

There’s one “best overall” in each answer, and it usually goes to the brand everyone already knows. But there are two or three specific need labels in the same answer, and those are much more winnable. A smaller brand that’s clearly the best option for one kind of shopper has a real path into the answer.

Labels also show what AI understood about a product. If ChatGPT calls your moisturizer “best for oily skin” when you built it for dry skin, the information it found about you is unclear or contradictory. See what AI gets wrong about your products.

Where do the labels come from? Mostly from the sources AI cites: review sites and testing roundups that already sort products into “best for” groups, plus the product’s own pages. See the sources ChatGPT cites.

What to do with this

  1. Pick the need you want to own. One clear “best for” per product, chosen from what your customers actually say.
  2. Say it plainly on the product page. “Made for dogs over 50 lbs who pull” in the first lines, not buried in the description.
  3. Back it with specifics. Ingredients, materials, sizes, test results: the details that make the claim checkable.
  4. Ask reviewers to describe themselves. “I have oily skin and this didn’t break me out” helps AI match you to the next oily skin shopper.
  5. Get the same label from outside. Pitch the “best for” angle to the roundups and review sites in your space.
  6. Check what AI calls you now. Ask ChatGPT, Gemini and AI Mode for the best products in your space and note the label next to your name, if you’re there.

How AI describes your products, and how to change it, is part of my AI Search and Conversion Audit. The AI Visibility Check is a quick first look.

How this was done

  • Answers: the 48 short buyer questions from Who ChatGPT recommends, 8 in each of six product areas, asked in ChatGPT on November 4, 2026 and in Gemini and Google AI Mode on November 5, 2026, each in a fresh session.
  • Labels: phrases such as “best overall”, “best value”, “best for …”, “budget pick” and “top pick” in the answer text, grouped by type with keyword rules and checked by hand. Questions asking “which is best for you” were excluded.
  • Limits: 48 questions per AI. Labels depend on the wording of the question, so a question that names a need (“for sensitive skin”) gets more need based labels. See how question wording changes AI recommendations.

No store or brand is named in this report.

FAQ

Questions, answered.

Does ChatGPT label the products it recommends?

Almost always. In 48 shopping questions I asked in November 2026, ChatGPT gave at least one "best for" style label in 44 answers, 149 labels in total. Gemini labeled picks in 31 answers and Google AI Mode in 29.

What labels does AI give products?

Mostly specific needs. 45% of ChatGPT's labels and 51% of Google AI Mode's named a shopper situation, like "best for sensitive skin" or "best for heavy pullers". About a third said "best overall", and most of the rest were about value or budget.

How do I get AI to call my product best for something?

Make the claim easy to find and easy to verify. State clearly on the product page who the product is for, back it with specifics, and make sure reviews and outside sources describe the same use. AI tends to repeat labels that several sources agree on.

Is it better to be best overall or best for a specific need?

For most brands, a specific need. There's only one "best overall" per answer, usually a well known brand. Most labels were for specific needs, so there are more of those spots to win.

See where you stand in AI search.

Run the free AI Visibility Check. In about a minute you'll see whether ChatGPT recommends you, who it recommends instead, and what's holding you back.