How do you get ChatGPT to recommend your brand?

The first source is training. OpenAI says the models behind ChatGPT are built from three kinds of information: what is publicly available on the internet, what it accesses through partners, and what users, trainers and researchers provide. The public part, it says, may include “publicly available webpages, public forums, public blogs, public posts”. Its crawler for training is GPTBot. OpenAI publishes a knowledge cutoff for its models in its API documentation, and a model's training covers only what was published before its cutoff, on your site and on everyone else's.

The second is search. OpenAI says ChatGPT may search the web automatically “when your question would benefit from current information”. It sometimes partners with other search providers, and when it does it typically rewrites the question into one or more targeted queries that it sends them. The privacy policies it points to for those providers are Microsoft's and Shopify's, so what its search reads is not limited to what OpenAI's own crawler collected. OAI-SearchBot is the crawler OpenAI uses to surface websites in ChatGPT's search features, and an answer that used search may carry citations you can open.

The two are not separate channels. For shopping questions, OpenAI lists the model's own earlier response, generated before it considers any new search results, among the things ChatGPT weighs when choosing which products to show. So there, what the model already believes about your category is one of the inputs even when it searches.

For a brand, that makes the crawler setting and its own pages the quickest levers: search can pick up a change once it reads the page again, with no wait for a new model.

Access is the one OpenAI publishes instructions for. It recommends allowing OAI-SearchBot in robots.txt and allowing requests from its published IP ranges, and says search takes about a day to adjust after the file changes. A site that opts out can still appear as a navigational link, OpenAI adds, but not in the search answers themselves. Its help page also asks site owners to confirm that their host or content delivery network lets that traffic through, and its publisher guidance sets the same condition for site content to be included “in summaries and snippets in ChatGPT”. OAI-SearchBot and GPTBot are separate settings, so disallowing GPTBot tells OpenAI your pages should not be used for training without taking them out of search. OpenAI's agents for search, for requests a person makes, and for training, and what OpenAI says each is for:

• OAI-SearchBot. OpenAI publishes it as the crawler used to surface websites in ChatGPT's search features.
• ChatGPT-User. OpenAI publishes it as the agent for certain user actions in ChatGPT, and says it is not used to decide what appears in search. OpenAI says that because a person starts these requests, robots.txt rules may not apply.
• GPTBot. OpenAI publishes it as the crawler for content that may be used to train its foundation models.

A fourth, OAI-AdsBot, checks pages submitted as ads, and is covered with ads below. Which crawlers to allow for the other assistants, and why a rule at your network edge can refuse one your file allows, is its own answer.

Then the pages themselves, and here the advice is ours rather than OpenAI's. Buyers usually describe a situation rather than name a product: a small team, a budget, a named use. Answer that situation directly on the page, in a passage that still makes sense read on its own.

And the facts on them. For products, OpenAI says ChatGPT weighs structured data such as price and description from first-party and third-party providers, and that a change to a merchant's prices can take time to show. For any brand, its own pages are among what ChatGPT's search can read, so a wrong price, date or specification there can reach an answer. Where an answer and your page disagree, the fix can sit on either side. Which of the three to start with depends on where your brand is weakest, though a search crawler you refuse comes first whatever else is true, and that order has its own answer.

You can see which of those pages ChatGPT reads in your own category. Ask a buying question with Search switched on, open Sources, and note the domains: OpenAI says Sources, when available, lists the cited sources and other relevant links. Where the same domains keep coming back on questions a competitor wins and you do not, those domains are where the work goes. Why a competitor comes up first has its own answer.

Those pages sit at different distances from you, and the work changes with the distance:

• A listing you supply. A retailer page, a marketplace listing, a directory profile. You can influence it directly by correcting what you supply, though the page is not yours.
• An independent review or comparison. You can give the writer accurate facts, access to people who can answer questions and, where it fits, the product to try. What they conclude is theirs.
• A forum or community thread. You can answer the question openly, under the brand's own name. The opinion belongs to the people in the thread.

None of it moves on your schedule. A page about you changes when the people who publish it decide to change it.

OpenAI began testing ads in the US on 9 February 2026. They may appear for users on the Free and Go plans, and not on Plus, Pro, Business, Enterprise or Edu. They sit below the end of a response, labelled as sponsored and visually separated from it, and where several are eligible, OpenAI weighs relevance and advertiser bids to choose which comes first. The product results in its shopping answers, it says, are selected independently and are not ads.

So an ad buys a slot below an answer, shown to part of ChatGPT's audience. A brand the answer leaves out is still left out, with its ad underneath, so a report that counts ads and recommendations together overstates how often the brand is recommended.

How search results are ranked is OpenAI's call. It says ChatGPT ranks them on multiple factors intended to help people find relevant, reliable information, and that placement is not assured.

OpenAI says the same question may yield different answers, and the variation is large. In Delphi's study of answers collected up to 24 August 2026, an assistant asked the same question more than once inside one measurement named the identical set of companies 21.5% of the time. A before and after built from one answer each compares two draws, and it can show a gain or a loss that nothing caused. How much evidence a change needs has its own answer.

The free check by hand sets out which questions to ask and what to write down for each answer, and comparing a before with an after adds to what you record each time. Three things are specific to doing it in ChatGPT:

• Keep your own history out. OpenAI says ChatGPT may use saved memories when it rewrites a search query. Start a temporary chat and choose Unpersonalized before you send the first message: OpenAI says that option does not use memory, custom instructions or plugins.
• Ask from the same place. OpenAI says ChatGPT may use an approximate location from your IP address for local results, and that a VPN or network location may affect it. Ask the before and the after from the same place and network.
• Allow for the clock. A robots.txt change first needs the adjustment OpenAI describes, set out above. A rewritten page can only show up once search reads it again, a third-party page once its publisher changes it, and a change in what the model learned only once a model with a later knowledge cutoff replaces it.

Traffic shows only part of it. OpenAI says ChatGPT adds utm_source=chatgpt.com to the links it sends from search results, so those visits can be counted. An answer that recommends you and sends nobody leaves no visit at all.

The study publishes that figure with the counts behind it and how each was made. A brand's standing in ChatGPT is one part of its AI visibility. Some of the work itself carries across and some does not. Accurate pages and third-party coverage are there for any assistant that reads the web. Crawler access is set operator by operator: most publish their own crawler, so allowing OAI-SearchBot says nothing about the rest.

The same question has its own answer for Gemini, Claude, Perplexity and Grok.

Delphi measures them side by side. Each time it measures a brand, it puts the same questions to ChatGPT, Gemini, Claude, Perplexity and Grok more than once and keeps every answer word for word. Its score, the Delphi Index, carries a margin set by how far the assistants disagreed with each other on your questions, and the dashboard and the history page do not report a change in the Index smaller than that margin as a change. Each assistant's share of answers naming you is shown beside the others' from one measurement to the next, so a change in ChatGPT alone is visible, though that figure carries no margin of its own.

Delphi asks each assistant through its API, the route developers use, rather than through the consumer app. Most of its questions reach ChatGPT without a live web search, so a page or crawler change shows in a Delphi reading first where an assistant searches as it answers, and elsewhere only once a newer model has learned it. Each paid level includes one or two measurements per brand a month. It is not a daily tracker, and an extra measurement costs $49 when a before and after needs one sooner.