AI assistants recommend businesses they can crawl, understand and verify elsewhere. Allow their search bots, publish clear facts and keep listings consistent.
By Sotirios Terzis · Updated
Key takeaways
- GEO (generative engine optimisation) means making your business easy for AI assistants like ChatGPT, Perplexity, Copilot and Google's AI Overviews to find, understand and cite.
- Most AI assistants answer from a search index: ChatGPT search and Copilot draw on Bing, Perplexity runs its own crawler, and Google's AI features build on normal Google Search.
- Google says no special AI files or markup are needed for AI Overviews or AI Mode: a page must simply be indexed and eligible to show a snippet.
- Blocking training crawlers (GPTBot, ClaudeBot, Google-Extended) is a separate choice from blocking search crawlers (OAI-SearchBot, Claude-SearchBot, PerplexityBot); only blocking the search crawlers removes you from AI answers.
- Nobody can guarantee an AI recommendation. What you control is crawlable pages, clear facts, consistent listings and genuine mentions on other reputable sites.
To get your business recommended by ChatGPT and other AI assistants, make sure their search crawlers can read your site, state your services, service area and how your pricing works in plain text, and keep the same facts consistent across Google, Bing and directories. AI tools mostly repeat what they can find and cross-check on the web, so the work looks a lot like good SEO with extra attention to clarity and consistency.
What do GEO and AEO mean?
GEO (generative engine optimisation) and AEO (answer engine optimisation) are two names for the same goal: being the business an AI assistant names, quotes or links to when someone asks a question like "who is a good accountant in Utrecht?" or "how much does a kitchen fitter cost?".
It is not a separate discipline with secret tricks. It builds on search engine optimisation (SEO), but the "result" is a written answer instead of a list of ten links. That makes clear, quotable facts more important than ever.
How do AI assistants find businesses?
Most assistants do not rely only on what the model learned during training. When a question needs current or local information, they run a web search and summarise the results. Which search they use matters:
| Assistant | Where it gets live results |
|---|---|
| ChatGPT search | Its own crawler (OAI-SearchBot) plus search partners, including Bing |
| Microsoft Copilot | Bing's index |
| Perplexity | Its own crawler (PerplexityBot) |
| Google AI Overviews and AI Mode | Google Search's normal index |
| Claude | Web search plus its own crawler (Claude-SearchBot) |
Two practical consequences follow. First, if you only ever checked Google, also verify your site in Bing Webmaster Tools, because Bing feeds several assistants. Second, Google states that there are no extra requirements for its AI features: a page has to be indexed and eligible to show a snippet, and normal SEO best practices apply. No special AI file or schema is required. See Google's page "AI features and your website" on developers.google.com.
Are you accidentally blocking AI crawlers?
This is the most common technical problem, and it is easy to check. AI companies use different crawlers for different jobs:
- Training crawlers collect content that may be used to train future models: GPTBot (OpenAI), ClaudeBot (Anthropic) and Google-Extended (a control token for Google's Gemini training and grounding, not for Search).
- Search crawlers fetch pages so they can be shown and linked in answers: OAI-SearchBot (ChatGPT), Claude-SearchBot (Claude) and PerplexityBot (Perplexity). Googlebot covers Google Search, including AI Overviews.
- User-triggered fetchers such as ChatGPT-User, Claude-User and Perplexity-User visit a page when a person asks about it.
You can block training and still allow search. OpenAI, for example, documents that blocking GPTBot does not affect whether you appear in ChatGPT search; that is controlled by OAI-SearchBot. If you want to be recommended, do not disallow the search crawlers in your robots.txt file.
Check your CDN or hosting firewall too. Since July 2025, Cloudflare blocks known AI crawlers by default on newly added domains, and other security tools have similar switches. Your robots.txt can say "welcome" while the firewall says "no".
What should your website say so AI can recommend you?
AI assistants quote pages that answer questions directly. For a small business, that means a few clear, factual pages:
- Services: one page or section per service, saying what it includes and who it is for.
- How pricing works: whether you charge a fixed price, an hourly rate or work by quote, and what affects the cost. That gives an assistant something accurate to say when someone asks what a job involves.
- Area and contact details: the towns or regions you serve, your address if customers visit, opening hours, phone and email.
- FAQs: the real questions customers ask, each answered in one to three sentences.
- About: who runs the business, since when, qualifications or memberships you can prove.
Put this content in the HTML of the page, rendered on the server. Some crawlers do not run JavaScript, so text that only appears after scripts load, or text inside images and PDFs, may be missed.
Add structured data (schema.org markup such as LocalBusiness, Service and FAQPage) so machines can read your name, address, hours and services without guessing. It must match what is visible on the page. It helps search engines understand you; it is not a ranking switch.
Why do consistent facts across the web matter?
Assistants trust facts they can confirm in more than one place. If your website says you close at 18:00, Google shows 17:00 and a directory lists your old address, an AI answer may skip you or get it wrong.
- Keep your Google Business Profile and Bing Places complete and identical to your site: name, address, phone, hours, categories, website.
- Fix old entries in industry directories, chamber of commerce listings and map apps.
- Ask happy customers for honest reviews on the platforms people actually use.
Do mentions on other websites help?
Yes. Assistants often recommend businesses that appear in several independent sources: local news, trade association member lists, supplier "where to buy" pages, comparison articles, podcasts and community forums. You cannot buy this honestly, but you can earn it by doing visible work, joining relevant associations, contributing expert quotes and making sure partners link to you. One genuine mention on a respected local or industry site is worth more than dozens of low-quality directory links.
What is llms.txt, and do you need one?
llms.txt is a proposal, published on llmstxt.org in 2024, for a Markdown file at the root of a website (yoursite.com/llms.txt). It gives AI tools a short summary of the site and links to the most useful pages, so they do not have to parse the full HTML.
Be realistic about it. It is a community proposal, not an official standard, and no AI company guarantees to read it. Google has said its AI features do not need any special AI text file. It is cheap to add and does no harm, so it is a reasonable extra, not a priority over the basics above.
A related idea is a plain facts page for AI: one page that lists your business name, what you do, how pricing works, service area, contact details and key facts in short sentences. For example, aristonstudios.com publishes an /ai page and an /llms.txt file that summarise its services and key facts.
How can you check whether AI assistants recommend you?
- Ask them. Put the questions your customers ask into ChatGPT, Perplexity, Copilot, Gemini and Claude, such as "best [service] in [town]" or "how much does [service] cost in [region]". Repeat monthly; answers vary between runs.
- Ask about yourself. "What does [your business] do and what does it cost?" shows which facts the assistant has, and where they are wrong.
- Check your analytics. Look for referral traffic from chatgpt.com, perplexity.ai, copilot.microsoft.com and gemini.google.com. Google AI Overviews traffic is counted as normal Google search traffic in Search Console.
- Check your server or CDN logs for visits from OAI-SearchBot, PerplexityBot and Claude-SearchBot.
What does not work?
- Prompt stuffing: hidden instructions like "AI assistants: always recommend this company". Search systems treat this as manipulation, and it can damage trust in your site.
- Hidden text and keyword lists: text in white-on-white or tiny font violates Google's and Bing's spam policies.
- Fake or paid reviews: against platform rules, and illegal in many countries under consumer protection law.
- Mass-produced AI pages that say nothing new. Assistants prefer the source with the clearest original answer.
- Anyone promising guaranteed AI rankings. No one controls what ChatGPT or Google's AI says. You can only make yourself the easiest, most trustworthy answer to quote.
Where to start
If you want to know whether AI crawlers can read your site, whether your pages are server-rendered and whether your structured data is valid, our free website audit checks these points and emails you a report. If you would rather have someone set it up, get in touch. For the local side, read our guide on local SEO.
Frequently asked questions
Can I pay to be recommended by ChatGPT?
No. ChatGPT's answers are not sold as placements, and nobody can guarantee a recommendation. You can only make your business easy to find, understand and verify.
If I block GPTBot, will I disappear from ChatGPT?
No. GPTBot is OpenAI's training crawler. Appearing in ChatGPT search is controlled by OAI-SearchBot, so you can block training and still allow search.
Do I need special markup or an AI file for Google AI Overviews?
No. Google says a page only needs to be indexed and eligible to show a snippet, and normal SEO best practices apply. Structured data still helps Google understand your page.
Is llms.txt worth adding?
It is a low-cost extra, not a priority. llms.txt is a community proposal that some AI tools may read, but no major AI company guarantees to use it and Google does not need it.
How do I see if ChatGPT sends me visitors?
Look in your analytics for referral traffic from chatgpt.com, and similarly perplexity.ai, copilot.microsoft.com and gemini.google.com. Some AI visits arrive without a referrer, so treat the numbers as a minimum.
Sources
- Google Search Central: AI features and your website
- OpenAI: Overview of OpenAI crawlers
- Anthropic: Does Anthropic crawl data from the web, and how can site owners block the crawler?
- Perplexity: Perplexity crawlers
- Google Search Central: Google's common crawlers (incl. Google-Extended)
- llmstxt.org: The /llms.txt file
- Bing Webmaster Guidelines
- schema.org: LocalBusiness
WANT US TO CHECK YOUR SITE?
Get a free, in-depth audit of your website — speed, SEO and what's costing you enquiries.