Part 01 — The engines disagree about ChatGPT's own feature.
On Nov 10, 2026 OpenAI launched ChatGPT Business Hub in beta — a dashboard where a business verifies its domain, claims its profile, manages its details and product feeds, and sees aggregated insights such as mentions and clicks. The obvious question every business owner is now asking: does verifying actually make ChatGPT recommend you more?
OpenAI's own help documentation answers this plainly: claiming a profile does not guarantee more visibility, higher placement or sales. So on day 6 we asked the engines themselves — the exact question, verbatim: "Does verifying my business on ChatGPT (Business Hub) improve my visibility in AI answers?"
"Yes, verifying your business on ChatGPT's Business Hub can enhance your visibility in AI responses… a higher likelihood of your business being featured in relevant AI-generated answers."
"No. Verifying or claiming your business in ChatGPT Business Hub does not directly improve its visibility, ranking, or recommendation frequency. OpenAI says it does not guarantee increased visibility…"
"…does not directly boost your algorithmic ranking or guarantee more frequent mentions in AI answers. According to OpenAI's documentation… a verified badge alone won't force the model to favor your business."
"Verifying your business on ChatGPT Business Hub does not improve your visibility in AI answers. OpenAI explicitly says claiming a listing doesn't guarantee more visibility… and advertising doesn't influence ChatGPT's independent answers."
Google returned no AI Overview for this question in our scan window.
Read the pattern, not just the verdicts. Every engine that retrieved sources said no. The one engine that answered from memory — six days after its own company shipped the feature — said yes, and described the benefit in exactly the promotional language OpenAI's documentation disavows. This is the same mechanism we measured in our retrieval study: answers built from cited sources behave differently from answers written from memory, and the uncited ones are where the confident errors live.
The companion question — "What is ChatGPT Business Hub and should my company claim its profile?" — showed the same split. Perplexity, Gemini, Claude and Google AI Overviews described the real feature accurately (profile claiming, product feeds, mention and click insights, launched in beta) from retrieved sources. ChatGPT, citing nothing, described a generic-sounding platform to "connect with customers" and listed "Enhanced Visibility" as the first benefit of claiming.
Part 02 — So should you verify? Yes — and we are testing it in public.
Our reading of the evidence: verify, but file it under accuracy and measurement, not ranking. You get control of your displayed details, and you get OpenAI's aggregated mention and click counts — a free, ChatGPT-only slice of exactly the thing we measure across five engines. What you do not get, by OpenAI's own account, is placement.
But "the docs say it doesn't help" is a claim, not a measurement — so we are measuring it. Ainswer completed domain verification and company identity review for its own Business Hub profile within the feature's first week. This scan is the baseline: 0 mentions in 138 answers, 0 citations of ainswer.app. We will keep scanning weekly with the same 30 questions and publish whether anything moves after the profile goes live — against the counterfactual every other brand in this table provides.
Part 03 — 60 days later: the category scoreboard moved. We didn't.
In September we published our first scan of this category — report #002 — including the row that embarrasses us. This month we re-ran the same 16 questions through the same four engines. Same method, 60 days apart:
| Recommended in | Perplexity | ChatGPT | Google AIO | Claude |
|---|---|---|---|---|
| Otterly — Sep | 40% | 6% | 40% | 35% |
| Otterly — Nov | 44% | 25% | 56% | 38% |
| Profound — Sep | 40% | 10% | 40% | 23% |
| Profound — Nov | 50% | 6% | 44% | 19% |
| Peec AI — Nov (new) | 44% | 12% | 22% | 25% |
| Ainswer — both | 0% | 0% | 0% | 0% |
Two stories in that table. Otterly grew — most sharply on ChatGPT, from 6% to 25% of answers. And Peec AI went from untracked to a quarter of all answers in the same window — a reminder that this category's shortlists are still soft enough for a newcomer to enter in weeks. The full November scan — 30 questions, five engines, 138 scored answers — looks like this:
| Brand | Share of answers | First brand named | Strongest engine | Weakest engine |
|---|---|---|---|---|
| Otterly | 51% | 31× | Gemini 73% | ChatGPT 27% |
| Semrush | 47% | 17× | Gemini 80% | ChatGPT 20% |
| Profound | 44% | 32× | Gemini 80% | ChatGPT 17% |
| Peec AI | 43% | 5× | Gemini 73% | ChatGPT 23% |
| Ahrefs | 31% | 13× | Gemini 47% | Claude 10% |
| Scrunch | 9% | 1× | — | — |
| Rankscale | 6% | 1× | — | — |
| Brand24 | 4% | 5× | — | — |
| Knowatoa | 1% | 0× | — | — |
| Ainswer | 0% | 0× | — | — |
Note the engine spread: the category's leaders are named in up to 80% of Gemini's answers and as little as 17–27% of ChatGPT's. A brand reading only its ChatGPT number would underestimate every competitor in this table by half.
Part 04 — Who feeds the answers: the most self-fed category we have measured.
Across the 138 answers the engines showed 1,146 citations. Here is the part that explains the scoreboard: 11 of the 20 most-cited domains are AI-visibility vendors' own websites. Not reviews of them — their own blogs and comparison pages.
| Most-cited domains | Citations | Who owns it |
|---|---|---|
| blog.hubspot.com | 33 | Neutral media / marketing blog |
| semrush.com | 33 | A tool in this table |
| youtube.com | 30 | Platform |
| tryprofound.com | 22 | A tool in this table |
| reddit.com | 21 | Community |
| seranking.com (incl. subdomain) | 21 | An SEO tool's own site |
| techradar.com | 14 | Media |
| otterly.ai | 13 | A tool in this table |
| orchly.ai / llmpulse.ai / siftly.ai / geoptie.com / dageno.ai / deepsmith.ai / geotoolbox.ai / kite.ai | 7–12 each | Eight more AEO vendors' own sites |
| ainswer.app | 0 | Us |
In every category we have studied — payments, coding assistants, Taiwanese consumer brands, B2B suppliers — vendor and third-party pages out-cite official sites. But this category is the extreme case: the companies that sell AI-visibility measurement are feeding the AI answers about AI-visibility measurement with their own published pages, and the engines are citing them. The brands that grew in Part 03 are the brands in this table's cited column.
Which brings us to the honest confession this page exists to correct. Since September we published four category studies — about payments, about coding tools, about skincare, about B2B. We never published one about our own category. The mechanism we kept documenting for other people's categories — publish the comparison content the engines retrieve, or someone else's page answers the question — was running against us the whole time, fed by the eleven vendor domains above. This report is us finally eating our own cooking — along with a set of measured comparison pages for this category: the ranked tools table, Otterly alternatives and Profound vs Otterly vs Peec. Whether it works will show up in the same weekly scans as the Business Hub experiment, and we will publish the number either way.
Method & limitations
- Scan: 30 buyer-style questions about AI-visibility / AEO tracking tools, United States market, English, run on Nov 16, 2026 through ChatGPT, Perplexity, Gemini, Claude and Google AI Overviews — 150 question × engine checks, each run multiple times with consensus. 138 returned answers; Google AI Overviews returned an answer on 18 of 30 questions.
- Counting: brand presence is exact name-token matching in the answer text, case-sensitive — no tone judgment involved in this report's percentages. "First brand named" is the first of the ten tracked brands to appear in an answer.
- Name collisions: one tracked product ("Mention") has a name that is an ordinary English word and was excluded as uncountable; "Profound" is counted case-sensitively, which can overcount when the adjective starts a sentence — treat its figures as an upper bound.
- 60-day comparison: the September baseline is the published table in report #002 (16 questions × 4 engines). The November re-run uses the same 16 questions and engines; Google AI Overviews answered 9 of the 16 this time, so its November percentages rest on fewer answers.
- Quotes: engine excerpts in Part 01 are from single representative runs of our scan; full texts and citations are in our raw data.
- Disclosure: Ainswer is a product in this category — this is a competitor's page, written by the vendor with the worst score in it. Facts about ChatGPT Business Hub come from OpenAI's public documentation and launch coverage as retrieved on Nov 16, 2026.
What is your brand's number?
Paste your site — one real buyer question across five engines, free, no email. If AI never names you, you'll see which sources feed the answers instead — the same analysis as this page, for your category.