How to Check If Your Company Shows Up in ChatGPT, Google AI
A step-by-step way to check whether ChatGPT and Google AI find, understand and cite your company, and what to fix if they don't.
A buyer doesn't always start with a list of blue links in Google anymore. They might ask ChatGPT to compare vendors, ask Google AI for the best option, or describe their problem without knowing any company or product names at all.
This isn't a distant scenario. The Anthropic Economic Index, June 2026 shows that conversations about starting a business peak on weekends. Research keeps happening even when your sales team is offline.
You can check AI visibility without building an analytics stack. You just need to measure more than whether your name shows up at all.
What "visible in AI" actually means
Visibility has three levels:
- AI finds you. Your name appears in an answer related to what you do.
- AI understands you. It describes your services, specialty, audience and differences from other options correctly.
- AI cites you as a source. The answer links to your site or references information published there.
A mention alone isn't enough. If the model gets your offer wrong, relies on an outdated description, or can't point to a source, you're present, but your visibility isn't doing any work for you yet.
An AI visibility audit, step by step
1. Write down the questions a buyer actually asks
Don't start with your company name. Someone who already knows it is at the end of the journey. The most valuable questions come earlier.
Prepare 10 to 20 questions across four groups:
- Problem: "How do I improve my store's visibility in Google and AI answers?"
- Category: "Who builds fast websites for B2B companies?"
- Comparison: "Webflow or custom code for a site with a CMS?"
- Verification: "What's known about company X and what have they built?"
Use language from sales calls, emails and client inquiries. Internal service names often differ from what people actually type and say.
2. Check answers in clean sessions
Ask the same questions in a fresh ChatGPT conversation and in Google, when an AI Overview or AI Mode appears for that query. A new session limits the influence of prior context.
For each answer, note:
- whether your company was mentioned,
- whether the description of your offer is accurate,
- whether it linked to the right page,
- which sources the answer draws on,
- which other companies get mentioned,
- what's missing compared to competing sources.
One query doesn't give you a reliable picture. AI answers can shift, so the pattern across your full question list matters more than any single result.
3. Check the data, not just the answers
A manual test shows you a result, not the scale of the problem. In Google Search Console, check queries, landing pages, impressions and clicks tied to the topics your buyers ask about. Google also offers visibility reports for generative features like AI Overviews and AI Mode. This data helps you tell a lack of demand apart from a visibility problem.
You can spot ChatGPT traffic in your analytics by referral source. OpenAI confirms that publishers allowing OAI-SearchBot can measure visits from ChatGPT in tools like Google Analytics.
Not every contact leaves an obvious trail, though. Someone might read an AI answer, remember your name, and come back directly later. So traffic, mentions and inbound inquiries need to be read together.
4. Make sure crawlers can actually reach your site
Content a crawler can't fetch has little chance of becoming an answer's source.
Check:
- that important pages don't carry a
noindextag, - that
robots.txtdoesn't block Googlebot or OAI-SearchBot, - that your server, CDN or firewall isn't returning 403s to crawlers,
- that every important page returns a 200,
- that key content exists as text in the HTML,
- that ordinary internal links lead to the page.
OpenAI recommends allowing OAI-SearchBot so content can be discovered, cited and linked in ChatGPT Search. That's a different goal from training data crawling, which a separate bot, GPTBot, handles.
For Google, the rule is simpler. Your page needs to be indexed and eligible to show with a snippet. Google doesn't require a special tag or a separate file just for AI Overviews.
5. Check whether your site actually answers clearly
A crawler can reach your site and still not find a good answer there. That happens when your offer is described in generalities, or key facts are scattered across too many pages.
A good page should clearly explain:
- what your company does,
- who it's for,
- what problem it solves,
- exactly what the service includes,
- who's responsible for delivery,
- what proof of work exists,
- what the next step looks like.
Descriptive headings, short answers to specific questions, case studies with real numbers, and consistent company information all help. Schema.org structured data can clarify meaning, but it has to match what a visitor actually sees on the page.
Why a company might not show up in answers
A missing mention isn't always one SEO mistake. The most common causes are more mundane:
- there's no dedicated page answering that specific question,
- the offer is written in company language, not the buyer's language,
- crawlers can't access the content,
- brand information is inconsistent between the site and external profiles,
- there's no verifiable, citable proof,
- the content is correct but too similar to dozens of other pages,
- the site is new and hasn't built topical recognition yet.
AI doesn't pick sources on a single switch. Visibility comes from a combination of technical access, solid SEO, clear content, and credible signals from beyond your own domain.
What to fix first
Order matters. Publishing more articles won't help if a crawler can't read your site, or your offer doesn't have its own indexable pages.
1. Access and indexing
Clear technical blockers first. Check indexing status, server responses, robots.txt, your sitemap, and internal linking.
2. Service pages that match intent
Every core service needs a page that answers buying-stage questions. One catch-all services list rarely gives a search engine or a model enough context.
3. Content grounded in real experience
Publish answers that can't be produced by rewriting the top Google results. Your own data, selection criteria, project outcomes and honestly stated limitations are more useful than another generic guide.
4. A consistent company description
Name, offer, author, contact details and external profiles should all describe the same company. That makes it easier to tie scattered information back to one recognizable entity.
5. Regular measurement
Repeat your question set once a month. Log the date, the answer, the sources, and where your company ranked in the response. After a few months you'll see a trend, not a random screenshot.
What isn't enough on its own
A few things get pitched as shortcuts to AI visibility. Each can help, but none replaces a good site and credible content.
- An
llms.txtfile organizes information for some systems, but doesn't guarantee citation. - Structured data helps describe content, but won't fix a vague offer.
- A large volume of AI-written articles builds no advantage if they all repeat the same information.
- Keyword repetition doesn't make an answer more useful.
- A single ChatGPT test isn't a visibility measurement.
Frequently asked questions
Does llms.txt improve visibility in ChatGPT?
It organizes information for some systems, but on its own it doesn't guarantee your company gets cited. It works alongside a technically accessible site and specific content, not instead of them.
How often should I check my company's AI visibility?
Once a month, on a fixed set of 10-20 questions. A single ChatGPT test gives you a random snapshot, not a trend.
A short checklist
Your company has the basics of AI visibility if:
- important pages are indexable and accessible to crawlers,
- every core service has its own specific description,
- content answers real buyer questions,
- your offer and company data are consistent across the internet,
- case studies show real facts and outcomes,
- AI answers link to the right pages,
- visibility gets measured on a fixed set of questions.
AI visibility starts with a site people can trust
AI search readiness isn't a separate marketing trick. It's the sum of what a good site should already be doing: accessible, fast, clear, specific and backed by evidence.
The best first step is a short audit. A dozen or so questions, Search Console data and a technical check are usually enough to see whether the problem sits in access, content or credibility.
If you want your site checked, I can run that audit and point to the changes with the biggest impact first.