How to get your business recommended by ChatGPT and Google’s AI answers

Nobody can guarantee that ChatGPT, Google’s AI Overviews or any other AI assistant will recommend your business. These systems decide for themselves, and they change often. What you can do is make it easy for them: let their crawlers reach your site, publish clear, factual pages in plain HTML that say exactly what you do, where and for whom, answer common questions directly, add structured data, and keep your business details consistent across the web. Most of this is the same work that helps ordinary search, which is why it’s worth doing either way.

Be wary of anyone selling guaranteed placement in AI answers. No one outside these companies controls what their models say, and the systems change faster than any guarantee could keep up with.

How do AI assistants decide what to recommend?

Broadly, in two ways. Some of what a model “knows” comes from the text it was trained on, gathered from the web well before you ask it anything. And many assistants now search the web live when answering, read a handful of pages, and summarise them, often with links. Google’s AI Overviews draw on Google’s own search index. ChatGPT search, Perplexity and Claude use their own crawlers and search providers.

For a local or specialist business, the live search part matters most. The assistant needs to find pages that clearly answer the question, from sources it considers reliable, and with facts that agree with each other. That means your own site, your Google Business Profile, directories, reviews and anything else written about you.

Make sure AI crawlers can actually read your site

Many AI crawlers fetch the raw HTML of a page and may not run JavaScript the way a browser does. If your services, prices or contact details only appear after scripts load, some crawlers may see an empty page. This is a common issue with sites built entirely as single-page apps, and with some AI-generated builds.

• View the page source (not the inspector) and check your main text is there in the HTML.
• Don’t put key information only inside images, PDFs, sliders or tabs that load on click.
• Make sure important pages aren’t blocked by robots.txt, a noindex tag or a login.
• Keep pages reasonably fast. Our guide to Core Web Vitals explains what to check.

Which AI crawlers should robots.txt allow?

Your robots.txt file tells crawlers what they may fetch. The major AI companies publish the names their crawlers use, and they separate training from search, so you can choose:

• OpenAI: GPTBot collects content that may be used to train its models. OAI-SearchBot is used for ChatGPT’s search results. ChatGPT-User fetches pages when a user asks ChatGPT to.
• Anthropic: ClaudeBot collects content that may be used for training. Claude-SearchBot indexes for search results, and Claude-User fetches pages when a user asks Claude a question.
• Perplexity: PerplexityBot indexes sites for Perplexity’s search results. Perplexity-User handles user-requested fetches.
• Google: Google-Extended isn’t a separate crawler but a control token. It lets you opt out of your content being used for Gemini models, without affecting Google Search. AI Overviews are part of Google Search and are governed by normal Googlebot access.

If you want to appear in AI answers, don’t block the search crawlers. Whether you allow the training crawlers is a business choice: blocking them won’t remove you from search features, but allowing them may help models learn about you over time. Check the rules you already have, because some security plugins, CDNs and hosting providers now block AI crawlers by default. Crawler names and behaviours change, so check each company’s own documentation before editing the file.

Write pages an AI can quote

AI answers favour content that states facts plainly. A page that opens with “We’re passionate about delivering excellence” gives an assistant nothing to work with. One that says “We’re a licensed electrician in Geelong servicing the Bellarine Peninsula, specialising in switchboard upgrades and EV charger installation” does.

1. Give each service its own page, with a first paragraph that says what it is, who it’s for and where you offer it.
2. State concrete facts: service areas, opening hours, typical price ranges if you’re comfortable sharing them, licences, accreditations and memberships.
3. Add an FAQ section answering the questions customers actually ask, in their words. Keep each answer self-contained.
4. Use clear headings that match real questions, like “How much does a switchboard upgrade cost?”.
5. Keep an accurate About page with your business name, location, history and who runs it.
6. Keep pages current. Old prices, past staff and closed locations confuse people and machines alike.

Add structured data

Structured data, usually written in JSON-LD, describes your business in a form machines can read without guessing. Use LocalBusiness or a more specific type for your business details, Service for what you offer, and FAQPage where it suits a genuine FAQ. It doesn’t guarantee any particular treatment, and Google has reduced how often some types appear as special search features, but it removes ambiguity about the facts. Our guide to local SEO for service businesses explains the setup.

Keep your business facts consistent everywhere

When an assistant pulls information from several sources, conflicting details are a reason to trust you less, or to give a customer the wrong phone number. Make your name, address, phone, hours and service list match across your website, Google Business Profile, Apple Business Connect, Bing Places, industry directories and social profiles. Genuine reviews and mentions on reputable local sites also help, and that’s something no website change can replace.

Should I add an llms.txt file?

llms.txt is a proposed convention: a plain-text Markdown file at the root of your site that summarises what the site is and links to its most useful pages, written for AI tools. It’s cheap to add and does no harm. But it’s an emerging idea rather than an established standard, and the major AI companies haven’t committed to using it. Treat it as optional, and don’t expect it to change anything on its own. Clear pages and open crawler access matter far more.

Can I see whether it’s working?

Only partly. Ask the assistants the questions your customers would ask, from time to time, and note whether you appear and whether the details are right. Some AI tools pass a referrer when someone clicks through, so you may see visits from sources like chatgpt.com or perplexity.ai in your analytics; our comparison of GA4, Plausible and Fathom covers how to track that. Results vary between sessions and users, so look for patterns over months, not individual answers.

If you only do three things

Make sure search crawlers, AI and traditional, can reach plain-HTML pages on your site. Give each service a clear, factual page with real answers to real questions. Keep your business details identical everywhere they appear. That’s the honest core of it, and it helps ordinary search too.

If you’d like your site checked for crawler access, structured data and clear service pages, that’s a fixed-price technical SEO audit. If the site needs rebuilding into proper HTML pages, our small websites work covers that. Audits from $600. Most technical SEO jobs land between $600 and $3,000.