How to get your business named by ChatGPT, Gemini and Perplexity

To get your business named by ChatGPT, Gemini, and Perplexity, make your site easy to reach, easy to read, and easy to quote. Let AI crawlers in. Put plain answers about what you do, who it is for, and what it costs in text on the page. Then earn mentions on other trusted sites. Nobody can promise a mention.

What is the difference between SEO, AEO, and GEO?

They are three names for one job: help people find you. The difference is where they look.

Most of the work is the same for all three. A page that a search engine can read and trust is also a page an assistant can read and trust. The rest of this article shows the checks that matter and why.

If you want to see where your site stands first, the free search and AI visibility report runs these checks on your home page in a few seconds.

Do you need special AI files to show up in Google's AI answers?

No. Google says so directly. Its Search Central guide to AI features states that you do not need new machine-readable files, AI text files, or special schema to appear in AI Overviews or AI Mode. Normal SEO best practice applies.

That guide lists the basics: let Google crawl the site, link your pages to each other, keep important content in text form, and make structured data match the visible text.

Other assistants work differently. ChatGPT search, Perplexity, and Claude send their own crawlers to read pages. Some sites also add extra files such as llms.txt. So the basics come first, and the extras are cheap insurance.

Step 1: Let AI crawlers read your site

An assistant cannot quote a page it cannot fetch. Your `robots.txt` file tells each crawler what it may read.

Each company names its crawlers, and each crawler has a job:

• OpenAI uses GPTBot for model training and OAI-SearchBot for ChatGPT search. Its crawler documentation says sites that block OAI-SearchBot are not shown in ChatGPT search answers.
• Perplexity uses PerplexityBot to surface and link sites in its results, per its bot documentation.
• Anthropic uses ClaudeBot for model training, Claude-SearchBot for search, and Claude-User to fetch a page when a person asks Claude a question, per its crawler help article.
• Google uses Google-Extended to let sites limit AI training and grounding in some of its other systems. For AI Overviews and AI Mode, Google's guide points to normal Search crawling.

Check two more places. A firewall or bot-protection service can block crawlers even when `robots.txt` allows them. And a missing `robots.txt` is a missed chance to name your sitemap.

Decide on training separately from search. You can block training crawlers and still allow search crawlers. That is a business choice, so ask whoever owns your data policy.

Step 2: Put your answers in text that loads without JavaScript

Google's guide asks for important content in text form. If your home page draws its words with JavaScript, a crawler that does not run scripts sees a near-empty page.

Aim for a home page with at least 300 words of real text that load on their own. Say:

• what you do, in one plain sentence;
• who it is for;
• what it costs, or how you price;
• where you work, if location matters;
• how to reach a person.

Give the page one main heading (H1) that says what you do. Write a title of 10 to 70 characters and a meta description of 50 characters or more. Search results show these first.

Step 3: Answer the questions buyers ask

Assistants answer questions. They quote pages that answer the same questions plainly.

Write the questions your buyers type, word for word, as headings. Put a direct one- or two-sentence answer right under each one. Then add detail.

Good questions are specific: "How much does a bookkeeping cleanup cost?" or "Do you work with insurance agencies in more than one state?" Vague headings like "Our approach" give an assistant nothing to quote.

If you have a set of questions and answers, you can mark them up as `FAQPage` structured data. Keep the markup identical to the visible text.

Step 4: Describe your business with structured data

Structured data is a short block of code, usually JSON-LD, that states facts in a fixed format. It says "this is an organization, this is its name, these are its services, this is its address."

Google does not require it for AI features. It still helps any reader confirm basic facts without guessing. Use `Organization` or a more specific type such as `AccountingService` or `InsuranceAgency`. Add `Service` entries for what you sell. Never put a fact in the markup that is not on the page.

Step 5: Add llms.txt and a markdown version

These two files are optional. Google says you do not need them for AI Overviews. The llms.txt proposal is written for AI tools and agents, support varies by tool, and both files are small to set up.

• llms.txt is a short markdown file at the root of your site. The llms.txt proposal says it starts with a heading that names the site, then lists your key pages. Our guide what is llms.txt covers the format.
• A markdown version of a page is the same content as clean text, without menus and scripts. The proposal suggests a copy at the same address with `.md` added. You can also serve it when a tool asks for markdown, or link to it from the page.

Step 6: Fix the basics that search engines still check

Google says normal SEO best practice applies to its AI features. These basics are part of that:

1. Serve the site over HTTPS.
2. Add a mobile viewport tag so the page works on phones.
3. Name a canonical address on each page so duplicates do not split your ranking.
4. Publish a sitemap at `/sitemap.xml` or list it in `robots.txt`.
5. Add an Open Graph title and image so shared links show a preview.

Step 7: Make it easy to contact you

An assistant that names you should also be able to send the buyer somewhere. Link a contact page, a booking page, or an email address from your home page. A buyer who asks "who can help me with this?" needs a next step.

Step 8: Earn mentions on other sites

Your own site is one source. Assistants also read what other sites say about you. Reviews, industry directories, association listings, podcast notes, and local or trade press all count.

Keep your name, address, and description the same everywhere. Ask happy customers for reviews on the platforms your buyers use. Answer questions in public where your buyers ask them. This part is slow, and no file or tag replaces it.

What the free report checks

The free search and AI visibility report asks six short questions and checks your home page and a few files at your site root. It looks at:

• Search basics: HTTPS, title and description, one main heading, readable text without JavaScript, canonical link, mobile viewport, sitemap, and share preview.
• AI answers: whether `robots.txt` blocks GPTBot, ClaudeBot, PerplexityBot, or Google-Extended; JSON-LD that describes your business; FAQ data or question headings; llms.txt; a markdown version; an optional agent endpoint; and a contact or booking link.
• App stores: only if you have an app.

The crawler check reads `robots.txt` for those four names only. GPTBot, ClaudeBot, and Google-Extended are training crawlers, so the report does not check OAI-SearchBot or Claude-SearchBot, the crawlers behind ChatGPT and Claude search. Check those two yourself. If you block a training crawler on purpose, treat that line as a reminder of your choice, not a fault. If your firewall blocks the check itself, the report says so instead of scoring the site.

It scores each area, lists your top three fixes, and names one next step. You see it on the page and get it by email. It does not check third-party mentions, and it cannot tell you whether an assistant will name you. It tells you whether your site gets in the way.

Your next step

Most of these fixes are small jobs for whoever runs your website. If you want help deciding what to do first, or you want to see where AI workflows and agents fit in the rest of your operations, the $1,500 AI Assessment takes one week and ends in a ranked plan or a clear no. Or book a call and ask your question directly.