"What is the best accounting software for a small agency?" "Which running shoes are good for flat feet?" Buyers now put questions like these to ChatGPT, and they get a short list of named brands back.
If you are not on that list, you are not in the running. This guide explains how ChatGPT, Perplexity and Google's AI features choose what to recommend, and what you can do about it.
How does ChatGPT decide which brands to recommend?
For buying questions, ChatGPT usually searches the web, gathers candidate pages from its search providers and its own crawler, and builds an answer from sources that best match the query. Brands that appear consistently across trusted pages — reviews, comparisons, product data and their own site — are the ones most likely to be named.
There are two sources of knowledge at play. The first is what the model learned during training. The second is live retrieval: when ChatGPT searches, it pulls current pages and cites them.
For purchase-intent questions, live retrieval matters most. Prices, availability and new products change too fast for training data alone.
OpenAI is clear that no one can buy or guarantee a spot. Its publisher FAQ states there is no way to guarantee top placement in ChatGPT search, and that ranking depends on factors designed to surface reliable, relevant information.
For a broader playbook, see how to get your brand into ChatGPT and AI search. For the underlying discipline, read What is answer engine optimization?
Where does ChatGPT search get its sources?
ChatGPT search draws on third-party search providers, content from partners, and its own crawler, OAI-SearchBot. Each is a door your brand can come through.
OpenAI's help centre explains that ChatGPT may share disassociated search queries with Bing to return web results. That makes Bing indexation a practical requirement, not an afterthought. Submit your sitemap in Bing Webmaster Tools and use IndexNow to notify Bing when pages change.
OpenAI also runs its own crawler. According to its crawler documentation, OAI-SearchBot is used to surface websites in ChatGPT's search features. Sites that block it will not be shown in ChatGPT search answers, though they can still appear as navigational links.
Two separate bots are often confused with it:
- GPTBot collects data for training models. Blocking it does not remove you from search answers.
- ChatGPT-User fetches pages when a user's request triggers it. OpenAI notes robots.txt rules may not apply to these user-initiated actions.
How does ChatGPT pick products in shopping answers?
ChatGPT selects products based on how relevant they are to the user's request, using structured product data from merchants and third-party providers plus other web content such as reviews. OpenAI states these product results are not ads.
According to OpenAI's shopping help article, products appear when ChatGPT judges them relevant to the user's intent, taking into account the query and context such as Memory or custom instructions. Results are selected independently and are not influenced by OpenAI partnerships.
When several merchants sell the same product, the article says ChatGPT weighs factors such as availability, price, quality, whether the merchant is the primary seller, and whether Instant Checkout is enabled.
It may also show review summaries generated from reviews on public websites. And merchants can supply data directly: OpenAI's Agentic Commerce Protocol accepts structured product feeds with current price and availability.
For e-commerce brands, the takeaway is practical:
- Keep product titles, descriptions, prices and stock accurate everywhere.
- Explore a direct product feed if you sell online.
- Earn reviews on the platforms your buyers use.
How do Perplexity and Google AI Overviews choose sources?
Perplexity retrieves pages through its own crawler and cites them inline, while Google builds AI Overviews and AI Mode from its Search index using a query fan-out process. Both reward clear, relevant, accessible pages.
Perplexity's crawler documentation says PerplexityBot is designed to surface and link websites in Perplexity's results, and that it respects robots.txt. A separate agent, Perplexity-User, fetches pages when users ask questions and generally ignores robots.txt.
Google's AI Mode uses a query fan-out technique, splitting a question into subtopics and running many searches at once. A buyer's "best CRM for a small agency" may fan out into pricing, integrations, reviews and alternatives. Pages that answer those sub-questions can be pulled in.
To be eligible, Google says a page must be indexed and eligible to show with a snippet. For local and retail queries, Google also advises keeping Merchant Center and Business Profile information up to date. Our Google AI Overviews guide covers that surface in depth.
Why do third-party mentions, reviews and listicles matter so much?
AI engines answer "which brand?" questions by synthesizing what many sources say, so third-party mentions carry more weight than your own claims. A brand that only praises itself on its own site gives the model little to confirm.
Look at the pages that typically answer buying questions: "best of" lists, comparison articles, review platforms, forums and video reviews. Those are the pages retrieval systems find for queries like "top project management tools".
The correlation data supports this. An Ahrefs study of 75,000 brands found branded web mentions correlated with brand visibility in ChatGPT at about 0.66, while YouTube mentions reached about 0.74. Link-based metrics showed much weaker correlations. Correlation is not causation, but the pattern is consistent across ChatGPT, AI Mode and AI Overviews in that study.
Consistency matters too. When your name, category and key facts match across your website, directories and reviews, engines can connect the dots. Contradictions make you easier to skip. This off-site layer is what separates GEO from classic SEO, as our AEO vs SEO vs GEO comparison explains.
Which signals can a brand actually influence?
A brand can directly control its crawler access, site content and product data, and can influence reviews, listings and editorial coverage. The table maps each signal to where it lives and the action that helps.
| Signal | Where it lives | What to do |
|---|---|---|
| Crawler access | robots.txt, CDN, firewall | Allow OAI-SearchBot, PerplexityBot, Bingbot, Googlebot |
| Index presence | Bing, Google | Submit sitemaps; use IndexNow for updates |
| Clear product and service facts | Your website | Answer-first pages, comparison tables, specifics |
| Structured product data | Feeds, Merchant Center | Keep price, stock and descriptions current |
| Reviews | G2, Trustpilot, Google, marketplaces | Ask real customers consistently |
| Editorial mentions | Comparison articles, press, podcasts | Pitch data and expertise, not ads |
| Local data | Google Business Profile, directories | Keep name, address and hours consistent |
OpenAI's FAQ adds a technical detail many sites miss: your host or CDN must allow traffic from OpenAI's published search-bot IP addresses. Aggressive bot protection can block OAI-SearchBot even when robots.txt allows it.
What steps should a brand take to get recommended by ChatGPT?
Start with access and facts, then build third-party proof, then measure. This order fixes the problems that block you entirely before spending on slower authority work.
- Audit access. Check robots.txt, firewall and CDN rules for OAI-SearchBot, PerplexityBot, Bingbot and Googlebot.
- Confirm Bing and Google indexation of your key commercial pages.
- Write for buying questions. Create pages that answer "best X for Y", "X vs Y" and "X alternatives" honestly, including where you are not the best fit.
- State facts plainly. Who you serve, where, what you offer, and what makes you different, in visible text.
- Grow genuine reviews on the platforms your category uses.
- Earn inclusion in comparison articles by offering editors data, demos or expert commentary.
- Keep profiles consistent, including Google Business Profile for local businesses. Our local SEO team handles this side.
- Track results with a fixed prompt set (next section).
Avoid shortcuts. Google treats generating many pages primarily to manipulate rankings as spam, and fake reviews can damage trust with every engine.
How can you check whether ChatGPT recommends your brand?
Build a fixed list of buying prompts, run them regularly across ChatGPT, Perplexity and Google, and log whether you are named, cited and described correctly. Supplement that with referral and citation data.
Write 20–40 prompts that mirror real buyer language, such as "best [category] for [audience] in [location]". Run each several times, because answers vary between sessions. Record which brands appear, in what order, and which sources are cited.
Then add hard data:
- Referral traffic. OpenAI says ChatGPT adds utm_source=chatgpt.com to referral links, so you can segment those visits in analytics.
- Citation data. Bing Webmaster Tools' AI Performance report shows citations across Copilot, Bing AI summaries and select partners.
- Source analysis. Note which third-party pages keep being cited for your category. Those are your outreach targets.
For a quick first read on your site's AI readiness, try our free GEO checker.
Frequently asked questions
Can you pay to be recommended by ChatGPT?
No. OpenAI states product results in ChatGPT shopping are selected independently, are not ads, and are not influenced by OpenAI partnerships. Its publisher FAQ also says there is no way to guarantee top placement in ChatGPT search. Be sceptical of anyone selling guaranteed ChatGPT recommendations.
Does ChatGPT use Google or Bing?
OpenAI's help centre says ChatGPT search may share disassociated queries with Bing to return web results, alongside other third-party providers and partner content. OpenAI also runs its own OAI-SearchBot crawler. Being well indexed in Bing, and not blocking OAI-SearchBot, are therefore both important for ChatGPT visibility.
Will blocking GPTBot remove my brand from ChatGPT answers?
Not from search answers. GPTBot collects training data, while OAI-SearchBot surfaces sites in ChatGPT search. OpenAI's documentation treats them as separate settings. Blocking OAI-SearchBot is what removes you from ChatGPT search answers, apart from navigational links. Check both rules in robots.txt deliberately.
Why does ChatGPT recommend my competitor and not me?
Usually because your competitor appears in more of the pages ChatGPT retrieves for that question: comparison articles, reviews, forums or video. It may also have clearer facts on its site. Run the prompt, look at the cited sources, and see where your brand is missing or described inaccurately.
Do reviews affect ChatGPT recommendations?
They can. OpenAI's shopping help article says ChatGPT may show model-generated review summaries based on reviews from public websites. Review platforms also rank for many "best X" queries that feed retrieval. Genuine, recent reviews on the platforms your buyers use are one of the most reliable signals a brand can build.
How long does it take to start appearing in ChatGPT recommendations?
Access fixes can work quickly: OpenAI says robots.txt changes are picked up in about 24 hours, and indexing follows recrawling. Earning reviews and editorial mentions takes months. Expect gradual gains in how often you are named, measured over several weeks, rather than a sudden switch.
If you would like to see how ChatGPT and other engines currently describe your brand, Meek Media offers a free audit that runs real buying prompts, checks crawler access and maps the sources your competitors win on. Our ChatGPT optimization and GEO teams can take it from there, or you can simply contact us with a question.