ChatGPT Business Hub Counts Your Clicks; 36 of 479 Top Sites Block the Bot Behind Them

OpenAI opened a beta on 10 November that lets businesses claim ChatGPT profiles and see 28 days of mentions and website clicks. Those numbers depend on OAI-SearchBot, which our crawl found fully blocked by 36 of 479 top sites with a readable robots.txt, including Amazon and Yelp.

Vibhu Dhariwal
Cofounder & CTO, Digital Gratified
Published
Section
Industry news
Reading time
6 minutes
Data card: 36 of 479 top sites with a readable robots.txt block OAI-SearchBot

OpenAI on 10 November 2026 opened a beta that lets businesses claim their profile in ChatGPT and see how often the assistant mentions them and sends clicks to their website, VentureBeat reported. The dashboard, called Business Hub, is the first first-party view of ChatGPT referral data most sites will get. Its numbers depend on a crawler some large sites still shut out: in our check of the 1,000 most popular domains, 36 of the 479 with a readable robots.txt fully block OAI-SearchBot, the agent OpenAI uses to surface sites in ChatGPT search.

What changed

  • Businesses can claim a ChatGPT profile, edit the details customers see and view 28 days of aggregate activity: mentions, website clicks and recurring discovery topics.
  • For an online business, a mention counts only when ChatGPT references its website; for a local business, showing the location is enough. Metrics appear only after they pass minimum thresholds.
  • Eligible US online retailers and service companies can also publish a Business Agent that answers customer questions inside ChatGPT, using information and guidelines the business supplies.
  • Claiming and the agent are free at launch, according to VentureBeat, which cited Digiday.

How Business Hub works

Verification can run through a business phone number, email address, website domain or Shopify account, VentureBeat reported. After approval, whoever manages the listing must pass an identity check through Persona before publishing changes, viewing analytics or creating an agent; a blue check appears once that is done. Local stores and restaurants can claim only an existing US listing, while claims for other businesses are open in markets where ChatGPT operates and OpenAI supports verification.

Retailers can upload a product feed or connect a Shopify store to keep prices and stock current. OpenAI said, according to VentureBeat, that it does not share user conversations or Business Agent chats with businesses, and that claiming a profile does not guarantee better placement. A screenshot published by Search Engine Roundtable shows the clicks metric carries a tooltip noting that recorded clicks do not confirm visits or conversions.

QuestionChatGPT Business HubGoogle Search Console
Main traffic metricRecorded clicks on links to your websiteClicks from Google Search results
Visibility metricMentions (website must be referenced for online businesses)Impressions and average position
Date rangePast 28 days, aggregateLast three months by default, adjustable with the date filter
Query detailRecurring discovery topics, not individual promptsIndividual search queries

The Search Console column reflects Google's Performance report documentation; the Business Hub column reflects VentureBeat's account of OpenAI's setup guide.

The crawler behind the clicks

OpenAI runs three crawlers with different jobs. Its crawler documentation says GPTBot gathers content for training foundation models, OAI-SearchBot is used to surface websites in ChatGPT search, and ChatGPT-User handles actions users start themselves. Sites that opt out of OAI-SearchBot "will not be shown in ChatGPT search answers", the page says, though they can still appear as navigational links. OpenAI adds that its systems can take about 24 hours to pick up a robots.txt change.

That makes robots.txt the first thing to check before reading anything into a blank Business Hub report. Our crawl on 13 November sorted the 479 top domains with a readable file by how they treat the three OpenAI agents:

Stacked bar of 479 top sites: 387 block no OpenAI crawler, 52 block GPTBot but allow OAI-SearchBot, 4 block only ChatGPT-User, 36 block OAI-SearchBot
OpenAI crawler rules in robots.txt across 479 of the Tranco top 1,000 domains that served a readable file. Source: SaaS Link Building crawl, 13 November 2026.

The 36 sites that block OAI-SearchBot include 11 Amazon country domains, Yelp, TikTok, Quora, Figma and a run of news publishers such as The New York Times, the BBC, CNN, Bloomberg and NPR. All but one, Weibo, also block GPTBot. The 52 sites that block GPTBot but leave OAI-SearchBot open are making the split OpenAI's documentation describes: no training, but eligibility for ChatGPT search. Another 50 sites name OAI-SearchBot in their file without blocking it, a sign that many larger publishers now manage OpenAI's agents one by one.

Yelp is the notable name for a product aimed partly at local businesses. Its block does not stop a restaurant from claiming a ChatGPT profile, but it means ChatGPT search answers should not show Yelp pages themselves, which could leave more room for a business's own site and profile. This is our inference from OpenAI's documentation, not something OpenAI has said about Business Hub.

Context: ChatGPT attribution has been messy

Until now, most sites measured ChatGPT traffic through referrers and the utm_source tag ChatGPT appends to links, and that tag has proved unreliable. We reported in October that ChatGPT links lost UTM tags after GPT-6, shifting visits into direct traffic in one sample. A click count recorded on OpenAI's side gives site owners a second number to reconcile against analytics, though the dashboard's own tooltip warns that clicks do not confirm visits.

The launch also lands in a week when AI companies are paying more attention to who supplies their answers. Yahoo became the first licensee of the Internet Licensing Collective on 12 November, and our check of ILC supporter sites found most still block AI crawlers outright. Several of the publishers that block OAI-SearchBot in our top-1,000 crawl, including USA Today, CNET and HuffPost, are listed as ILC supporters.

What SEO teams should do

  • Claim the profile for each brand or location you manage, and complete the Persona check early, since analytics stay locked until it is approved.
  • Read your robots.txt before reading the dashboard. If OAI-SearchBot is disallowed, ChatGPT search answers should not show your pages, and website mentions and clicks will be thin. Allow it if you want ChatGPT search visibility, even if you keep GPTBot blocked.
  • Check that your CDN or bot manager is not challenging OAI-SearchBot even when robots.txt allows it; verify the agent against the IP list OpenAI publishes for it (linked from its crawler documentation) rather than the user-agent string alone.
  • Record Business Hub clicks alongside analytics sessions from chatgpt.com each month, and expect a gap, since a recorded click is not a confirmed visit.
  • Use discovery topics as a content input. They show the themes in which ChatGPT cites your site, which is closer to intent data than a raw mention count.

Methodology

On 13 November 2026 we downloaded the Tranco top 1 million list and requested https://domain/robots.txt for the top 1,000 domains, following redirects with a 12 second timeout. 479 returned HTTP 200 with at least one User-agent line. An agent counted as blocked when a group naming it contained Disallow: / without an allow-all rule; wildcard groups were not counted, so sites that block every bot with a single User-agent: * rule fall outside these counts. Business Hub details come from VentureBeat's report and Search Engine Roundtable's screenshots, as OpenAI's own pages did not load for our tools.

Conclusion

Business Hub gives site owners their first official count of ChatGPT mentions and clicks, but the count can only reflect what ChatGPT is allowed to see. Most top sites leave OpenAI's search crawler open; the 36 that do not, including Amazon and Yelp, have chosen to stay out of ChatGPT search answers. For everyone else, the useful step this week is to claim the profile, confirm OAI-SearchBot access and start a baseline before the beta becomes a fixture.

Questions and answers

What is ChatGPT Business Hub?

Business Hub is an OpenAI beta, launched on 10 November 2026, that lets businesses claim their ChatGPT profile, edit their details and see 28 days of aggregate mentions, website clicks and discovery topics.

Does Business Hub cost anything?

Claiming a profile and publishing a Business Agent are free at launch, according to VentureBeat, which cited Digiday. OpenAI has left open the possibility of future charges.

What counts as a mention in Business Hub?

For an online business, ChatGPT must reference the business's website for a mention to count. For a local business, showing the location is enough. Metrics appear only after they pass minimum thresholds.

What happens if my site blocks OAI-SearchBot?

OpenAI's crawler documentation says sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers, though they can still appear as navigational links. Robots.txt changes can take about 24 hours to take effect.

How many top sites block OAI-SearchBot?

In our 13 November 2026 crawl of the Tranco top 1,000 domains, 36 of the 479 that served a readable robots.txt fully blocked OAI-SearchBot, while 52 blocked GPTBot but left OAI-SearchBot open.

More releases

  1. Industry news

    Brave Copies Google's /goto Link Cloaking; Our Pull Finds Google Now Hides All 164 Links

    Brave Search said on 11 November it will route result clicks through encrypted search.brave.com/go/ links after scrapers pushed out by Google's /goto redirects flooded its servers. Our pull of 20 Google result pages on 13 November found all 164 page-one organic links wrapped in /goto, up from 74% on 31 October.

  2. Studies

    ILC Launches With Yahoo Scout; Only 4 of 73 Supporter Sites Post an RSL License

    The Internet Licensing Collective launched on 12 November with Yahoo Scout as its first paying licensee. Our robots.txt check of 81 sites the ILC lists as supporters found only 4 of 73 readable files carry an RSL License line, while most fully block AI crawlers such as CCBot and ClaudeBot.

  3. Studies

    Do Title Tag Keywords Still Rank? Only 27.5% of 262 Google Titles Held the Exact Query

    Google says titles are a small ranking factor, but our pull of 262 US Google results for 30 queries found only 27.5% of title links held the exact query, and first place rarely had one. DOJ notes and leaked Google documents explain why titles still count.

Every release

Tell us the site and the goals.

Rankings in Google and visibility in AI answers, from one monthly plan. One call is enough to say which one fits. Email works just as well.