Why does ChatGPT not mention my company at all?
Why does ChatGPT not mention my company at all?
An AI assistant leaves a company out when it cannot read the company's pages, cannot tell what the company does, or finds no sources that connect the company with the question. Check in that order. Access for the AI search crawlers comes first. A plain statement of what you do and for whom comes next. After that come consistent company facts and your presence on the pages assistants cite. One day of sampled answers proves little on its own, since answers change from one run to the next.
What is happening
You ask ChatGPT, Perplexity or Claude for the best option in your category. The answer names four competitors and not you. Sometimes the assistant describes your company correctly when you ask about it by name, and still never recommends it. Both cases have causes you can check. We check them in a fixed order.
Why it can happen
AI search crawlers cannot fetch the pages
Each assistant uses a search crawler that is separate from its training crawler. OpenAI documents OAI-SearchBot for ChatGPT search and GPTBot for training. OpenAI says a site that blocks OAI-SearchBot will not be shown in ChatGPT search answers. It can still appear as a navigational link. Perplexity asks sites to allow PerplexityBot. Anthropic says that disabling Claude-SearchBot may reduce a site's visibility in Claude's search results.
A robots.txt rule or a firewall setting written to stop AI training can stop the search crawlers as well. We treat crawler access the way a security team treats account access. Each crawler gets its own decision, tied to its purpose, and someone owns that decision in writing. Blocking training and allowing search is a legitimate choice. Blocking both by accident is the one to rule out first.
Google adds one more switch. Since August 31, 2026, every Search Console property has a Search generative AI control. A property set to Exclude gets no traffic or impressions from AI Overviews and AI Mode. A child property can inherit that setting from its parent. Google says the control does not affect AI training, which Google-Extended controls, together with grounding in Gemini Apps.
The site does not say plainly what the company does and for whom
An assistant can only repeat what a page states. A homepage that says "we help brands grow" gives it nothing to repeat. Compare a sentence such as "Acme repairs commercial refrigeration in Phoenix for restaurants and grocery stores" (a made up example). That one sentence names a company, a service, a place and a customer. Our AI Legibility Framework calls this AI Readability: whether the key facts are stated plainly or buried in marketing prose.
Company facts conflict across the site and profiles
One version of this problem sits between a brand and the company that operates it. The markup folds the company into the brand, and the legal pages word the relationship differently. An assistant that reads both has two versions of one company to choose from.
Our own markup gives Amaretta LLC, Visibility Mesh, the web product and the Shopify app one entry each, with one name per entry. Our privacy policy names Amaretta LLC as the company that operates Visibility Mesh, and so do our terms of service, legal notice and contact information page. A check runs every day to confirm that every ID in our markup still carries exactly one name. For your own site, settle on one spelling for the legal name, the brand name and each product. Use it on your site, in your markup and on every profile you control.
The sources assistants cite for the category leave the company out
Assistants that search the web cite pages. They cite comparison articles, directories, review sites and trade press. When those pages list your competitors and not you, the answer follows them. Google says its AI features are rooted in its core Search ranking and quality systems. The pages that rank for your category's questions therefore tend to shape the answer too.
| Cause | First check | Source | What the check does not prove |
|---|---|---|---|
| AI search crawlers cannot fetch the pages | Read robots.txt for OAI-SearchBot, PerplexityBot and Claude-SearchBot, read the firewall events, and check the Search generative AI control | OpenAI, Perplexity and Anthropic crawler pages, and Search Console Help | A blocked crawler does not prove you are absent. Answers can cite other pages about you, or older data. |
| The site does not say plainly what the company does | Read the homepage and About page as a stranger would, and write down what they say you do, for whom and where | Visibility Mesh AI Legibility Framework v1.0 | Clear pages give an assistant something to use. The assistant still decides what to name. |
| Company facts conflict across the site and profiles | Compare the name, description, address and profiles on your site, in your markup and on each profile | Google Search Central, Organization structured data | Conflicting facts do not prove that any assistant confused you. Only a sample of answers shows what they say. |
| The sources assistants cite leave the company out | Ask several assistants the same buyer questions on several days and list the pages they cite | Google Search Central, AI optimization guide | A sample of answers is a snapshot of systems that change on their own. It is never a census. |
How to check which cause applies
- Check robots.txt, the firewall and the Search generative AI control for the AI search crawlers.
- Read the homepage and About page as a stranger would.
- Compare company facts across your site, your markup and your profiles.
- Sample answers to buyer questions several times.
Check robots.txt and the CDN for OAI-SearchBot, PerplexityBot and Claude-SearchBot
Open yoursite.com/robots.txt. Look for a group that names OAI-SearchBot, PerplexityBot or Claude-SearchBot. Look for a group for all user agents too. A line that says Disallow: / in the group that applies to a crawler blocks that crawler from the whole site. Your CDN or firewall comes next. A service such as Cloudflare can block AI crawlers even when robots.txt allows them. Read the firewall events for requests from those crawlers. OpenAI says a robots.txt change takes about 24 hours to reach its systems.
In Search Console, open Settings and read the Search generative AI control for the domain property. Read it for any child property too, since a child can inherit Exclude from its parent.
Read the homepage and About page as a stranger would
Give both pages to someone who does not know the business. Ask them to write one sentence on what you do, for whom and where. If they cannot, an assistant has little to work with.
Compare company facts across profiles
List the company name, legal name, description, address, phone and founder. Write down how each one appears on your website and in its Organization markup. Do the same for LinkedIn, your Google Business Profile and the directories of your field. Every difference is a question the assistant has to settle alone. Most of them take minutes to fix.
Sample answers for buyer questions several times
Write five questions a buyer would ask before choosing a provider like you. Add one question that names your company. Ask them in ChatGPT, Perplexity and Claude on three different days. Save each answer with its date and the pages it cites. One run proves little, since answers change between runs. A pattern across days and assistants is evidence. Google says no third party tool has access to its internal ranking or AI systems, and that includes ours.
What it affects
Buyers who shortlist options in an assistant never see your name. The clicks those answers send go to the sites they cite. Nobody reports a recommendation they never saw. The loss reaches your analytics only as fewer new inquiries, with no obvious cause.
What you can check and fix yourself
- Allow OAI-SearchBot, PerplexityBot and Claude-SearchBot in robots.txt if you want to appear in AI search. Decide separately about training crawlers such as GPTBot.
- Set the Search generative AI control to Include, or confirm that it inherits Include.
- Rewrite the first paragraph of your homepage and About page. Each should say what you do, for whom and where.
- Use one spelling of the company name everywhere.
- Put the same description on LinkedIn, your Google Business Profile and the directories you control.
- Ask real customers for reviews on the platforms your market uses.
If an assistant already says something wrong about you, fixing wrong AI answers about a store covers the corrections. The AI citation readiness checklist lists the checks page by page. For how ChatGPT search finds and cites pages, read how to get your website to show up in ChatGPT.
What needs a developer, a writer or outside authority
A developer changes firewall and CDN rules. A developer also edits robots.txt where the platform generates it, and the Organization markup in the theme. A writer turns vague homepage copy into plain statements. The same writer produces the pages that answer buyer questions from what the business knows first hand. Outside authority means earning a place on the pages assistants already cite: directories, comparison articles, trade press and reviews. It is the slowest part of the work, and buying it in bulk breaks Google's spam policies.
How Visibility Mesh finds it and what it can implement
What the scan detects
Every scan, the free one included, reads each page as it is served. It checks eight AI crawler names against the robots.txt rule for that page, among them OAI-SearchBot, Claude-SearchBot and PerplexityBot. It reports Google-Extended on its own as a control for Gemini training and grounding. Google documents that Google-Extended has no crawler of its own. The scan also scores whether the brand, its products and their attributes are stated the same way in the data AI systems read.
The Shopify app adds AI Mentions. It runs real buyer questions and shows who the assistants name, you or a competitor. It does not show which pages an answer cites. The Search generative AI control, your outside profiles and the pages assistants cite need a person to read them. We do that in a diagnosis.
What Visibility Mesh can implement, and on which platforms
- On Shopify, the app's Growth plan applies fixes after you approve each one, and you can undo them. It writes FAQ pages, product copy and SEO fields. It cannot edit robots.txt or theme files, so it gives you the steps for those.
- Our AI Visibility Setup for Shopify is done for you in a duplicated theme, in English, and sold by quote. Nothing goes live until you approve it.
- On Shopify and WordPress we can also make the changes ourselves and manage the work through Monthly AI Visibility Management, which is sold by quote.
- Your developer makes the changes on Webflow, custom sites and every other platform. We give the ordered fix plan, and a rescan shows what changed.
We fixed our own entity markup by hand, and a daily check keeps it fixed.
How each fix is validated
A robots.txt change is confirmed by fetching the file and running a rescan. A firewall change is confirmed in the firewall events, where real crawler requests should now get through. Markup is confirmed in the Schema Markup Validator. Presence in answers is confirmed only by asking the same questions again on fixed dates. Even then, a change is associated with the work and not proven to be caused by it.
How progress is measured
The dated baseline
Before anything changes, we record a baseline with its dates.
- We save the crawler access result and the Search generative AI setting.
- We save the answers to the fixed questions, with the pages they cite.
- We export the Search Console Generative AI performance report, which Google rolled out to all sites on August 31, 2026.
- We export GA4 sessions in the AI Assistant channel, filtered to your own hostname.
What is compared each month
In Monthly AI Visibility Management, the AI mention check runs every month on the questions we agreed with you, with several runs per question, and the results are kept as a trend. We also rescan the website at the same depth each month, and Margareta approves each monthly performance report before it is sent. Assistants change on their own, so we describe any change as associated with the work.
Questions people ask
How do small businesses show up in ChatGPT or AI search at all?
Small businesses show up when the sources an assistant reads describe them clearly. Size does not decide how clearly a business can be described. Let the AI search crawlers read your site. Make your homepage and About page say plainly what you do, for whom and where. Keep the same name, address and description everywhere you appear. Ask real customers for reviews, and answer the questions customers ask you most on your own pages.
Why does ChatGPT know my product exists but never recommend it?
Knowing a product and recommending it are different steps. When you ask about your company by name, an assistant can describe what it read about you. A request for the best option in a category works differently. The assistant picks the products its sources tie most closely to that use. Your pages may never state the use cases or the type of buyer. Third party pages may seldom list you. A product that nobody outside the company lists is rarely the one an assistant picks. Publish those facts plainly, and get listed where your category is compared.
Next step
If you would rather work through this with someone, tell us in your own words what is not working. A person replies within one business day, and we start from your scan and your own description. The troubleshooting guide for websites that search engines and AI assistants are not finding covers the same checks. How to check if AI can read your website covers the crawler side. For the wider picture, read how to grow traffic from search and AI assistants.
Describe the problem in your own words, and a person will reply within one business day.
Sources
We read every source below on September 26, 2026.
- OpenAI describes OAI-SearchBot, GPTBot and the 24 hour delay in its overview of OpenAI crawlers.
- Perplexity describes PerplexityBot on its Perplexity crawlers page.
- Anthropic describes Claude-SearchBot in its help article on crawling and blocking.
- Search Console Help explains the Search generative AI control, rolled out to all sites on August 31, 2026.
- Search Console Help explains the Generative AI performance report.
- Google's AI optimization guide was updated July 10, 2026. It covers core ranking systems and third party tools.
- Google's page on Organization structured data was updated September 8, 2026.
- Google's spam policies cover link spam. Google updated them August 28, 2026.
- Analytics Help describes the AI Assistant channel in GA4.
- Our AI Legibility Framework v1.0 defines AI Readability.