How do ChatGPT, Gemini, Perplexity and Claude decide which brands to recommend?
Our answer
Assistants name brands their web search finds and the wider web already discusses. Let their search crawlers in, keep SEO basics and state plainly what you sell. Then earn mentions on review and comparison pages.
- Confidence
- Medium
- Strongest proof
- 4 controlled tests
- 6 official rules, 4 datasets, 2 practitioners, 2 opinions or claims
- Sources
- 16 sources
- 18 statements
- O
GPAEM
New here? Take the 1-minute tour
Why this is the answer
3 reasons and our judgmentLetting assistants' search crawlers read your site is the entry ticket. Google's AI Overviews and AI Mode run on its core Search ranking systems (Google). OpenAI, Perplexity and Anthropic each say blocking their search crawlers can keep a site out of their search answers.
Brands the web already talks about get named more. Testing six AI models across five product categories, Edward Malthouse and co-authors found prominence tracks a brand's search interest and online conversation. Semrush credits Patagonia's steady visibility to outdoor review sites.
A brand's own pages matter most while the web doesn't know it yet. Answers citing Ahrefs' pages held 82% of its new brand's new mentions, against 6% for its established brand. Yet 43% of answers citing its conference pages left the new brand out.
Our judgment
- Treat search access and SEO basics as the entry ticket for every assistant. Google ties its AI features to core Search ranking, and OpenAI, Perplexity and Anthropic each tie search visibility to their crawlers.
- Claim your business in ChatGPT where it is open to you, and keep its information current. Treat it as upkeep, not a proven ranking lever, because OpenAI's post says nothing about recommendations.
- Put most of the effort into being discussed on third-party review, comparison and community pages. Malthouse and co-authors, Semrush's index and Jetfuel Agency's test all link recommendations to how widely the web discusses a brand, and third-party content drove most new mentions for Ahrefs' established brand.
- Onely's finding that ChatGPT cites brands' own pages first doesn't change where most of the effort should go. It covers 31 B2B agency questions, friction AI found brands' own sites were a small share of what assistants cited for consumer apparel brands, and Ahrefs found a brand's own pages mattered most while the web did not know it yet. It is a reason to make your own pages answer customers' questions plainly.
- Check visibility by asking your customers' questions several times over several days, not once. Ahrefs found about a quarter of citations appeared once and never again, Onely found answers change from run to run, and PPC Land reports Semrush's warning that model updates shift rankings between months. Opascope found the brand named first in a head-to-head question won more often, so ask comparisons in both orders.
Our judgment, built from Google, OpenAI, Perplexity, Anthropic, Edward Malthouse and co-authors, Semrush, Jetfuel Agency, Ahrefs, Onely, PPC Land, friction AI and Opascope: our reading of these sources, not a finding from any one of them.
What the sources say
18 statements from 16 sourcesOpenAI has opened Businesses in ChatGPT in beta: businesses can claim a listing, manage their information and see how people engage with them, and eligible ones can build an agent that answers customer questions.
“Claim your business, manage your information, and get insights on how people engage with you.”
An academic working paper finds AI search engines cite top-quintile brands 1.4x to 2.8x more than the average of 1,176 brands; in controlled tests, unfamiliar brand names ranked lower and content rewrites barely helped.
“Conversely, fact-preserving content rewrites yield negligible ranking gains.”
OpenAI says sites opted out of OAI-SearchBot won't appear in ChatGPT search answers, so allow it in robots.txt. GPTBot is a separate setting for training, and changes take about 24 hours.
“Sites that are opted out of OAI-SearchBot will not be shown in ChatGPT search answers, though can still appear as navigational links.”
Perplexity says to allow PerplexityBot in robots.txt and your firewall to appear in its search results; it is not used to train AI models, and its user-requested fetcher generally ignores robots.txt.
“Since a user requested the fetch, this fetcher generally ignores robots.txt rules.”
Anthropic says blocking Claude-SearchBot or Claude-User may reduce a site's visibility in Claude's search results; ClaudeBot is the separate training crawler.
“Disabling Claude-SearchBot on your site prevents our system from indexing your content for search optimization, which may reduce your site's visibility and accuracy in user search results.”
OpenAI says product feeds in ChatGPT are open only to approved partners, recommends concise, factual product copy, and says optional fields can improve answers but aren't required.
“Use concise, factual copy that helps users understand products.”
Opascope's test of 2,304 vote pairs found ChatGPT picked the first-named of two brands 54.7% of the time, about 63% for evenly rated brands, so brand tracking should ask in both orders.
Onely asked 31 agency questions of five AI engines: brands' own sites were 57% of ChatGPT's citations, best-of lists 22% and Reddit 1%, while lists were 59 to 77% for Perplexity, Gemini and AI Mode.
“Ask each question several times over several days, because the answers change from run to run”
This academic paper found six AI chatbots often left out established brands; prominence tracked search interest and online talk more than popularity, and specific needs changed which brands appeared.
“brands omitted from ordinary recommendations can remain conditionally retrievable when distinctive cues are supplied.”
In an illustrative test of electrolyte drink prompts, Jetfuel found ChatGPT named the same three or four brands, ones with years of brand search, editorial roundup mentions and many Reddit threads.
“Paid ads do not purchase AI citations.”
Writesonic, which sells AI visibility software, studied 4,670 brands from March to June 2026: citation share correlated 0.41 with visibility change, sentiment 0.04. It says this is associational, not causal.
“The analysis is associational, not causal.”
Ahrefs tracked 34 of its own pages across 9,886 AI answers: for a new brand, 82% of new mentions came from answers citing its pages; for its established brand, only 6%. No control group.
“self-promotional content seems to help most when your brand is a believable answer to the query already.”
PPC Land, on Semrush's study of 1,200+ brands, says AI mentions come from brand authority and category fit while citations come from content depth. ChatGPT cited 15.4 sources per answer, Gemini 3.3.
“Mentions come from brand authority and category fit. Citations come from content depth”
Semrush's index of 126 million US AI search prompts found only 36 brands held top-100 visibility on all four platforms every month, and that AI tools lean on reviews, retailers and publishers.
“AI platforms also rely on customer reviews, community discussions, independent publishers, retailers, and industry sources to understand and recommend brands.”
An academic test of three AI models found well-known skincare brands were recommended 100% of the time with equal specs, but a rival with a rating edge under 0.1 star broke that. Authority-style claims also did.
“well-known brands get recommended 100% of the time (IAI = 10.0) when all products have the same specifications”
Friction AI's test of 12 athletic apparel brands across five AI models found Knowledge Graph strength tracked recommendations loosely. Nike was in 71.8% of athleisure answers, New Balance 3.4%.
Unusual, an AI visibility software seller, ran 37,000 AI model searches. Category leaders showed up almost always but won only 25 to 41% of recommendation slots, and about half of niche brands never appeared.
Google says its AI search features run on its core ranking systems, so optimizing for them is still SEO. Pages must be indexed and snippet-eligible, and no special AI files or markup are needed.
“From Google Search's perspective, optimizing for generative AI search is optimizing for the search experience, and thus still SEO.”
In the full answer
For readers who sign in- What to do this week3 steps
- The numbers behind it3 figures from cited sources
- Where each source standsA chart placing each source on the question
- The strongest disagreementThe best case against the answer, from a cited source
- How sure we areWhat its confidence rests on, and what would change the answer