How Your Business Gets Mentioned in AI Search: A GEO Guide

There is no hidden setting or paid shortcut that makes ChatGPT, Gemini or Google's AI answers mention your business. Being mentioned is mostly the result of a good website, consistent business information and trustworthy sources that talk about you. Google's own guidance says the same: there are "no additional requirements" and no "special optimizations" needed to appear in AI Overviews. This guide explains how AI search works, what a small business can realistically do, and what is still unproven, without hiding the parts we do not know.

What are GEO and AEO?

GEO (Generative Engine Optimization) and AEO (Answer Engine Optimization) are names for work that tries to raise the chance that your content is used and mentioned in the answers of AI-powered search tools. Neither term has an official, universally accepted definition; the industry uses them in slightly different ways. We use them together to mean "showing up accurately and currently in AI answers."

By "AI search" we mean three things here: the AI Overviews and AI Mode in Google results, ChatGPT's features that search the web and cite sources, and answer-focused tools such as Perplexity. What they share is that the user sees a ready answer instead of a list of links, and sometimes source links appear beneath it.

How is it different from classic SEO?

In classic SEO the goal is to rank high in the results list for certain searches. In AI search the goal is to appear inside the answer or as a cited source. In practice the difference is smaller than you might think:

• The foundation is the same: Google says a page must be indexed and eligible to show in Search with a snippet to appear in AI Overviews. So basic SEO (a crawlable, indexed, useful site) is the floor for this too.
• The answer comes first: We do not know which part of a page answer-generating systems use, but clear, self-contained paragraphs make sense when the user is shown a ready answer.
• Clicks may fall: If the answer is given on the results page, the visitor may never reach your site. We come back to this in the measurement section.
• Being named matters: Even without a click, your brand being named may lead someone to search for you later. We cannot measure how often that happens.

How do AI systems choose their sources?

The honest answer: we do not fully know. The companies do not publish how they rank and select sources, and it changes over time. What we could verify:

• Google: AI Overviews and AI Mode have no separate "AI index" requirement; indexed pages that meet Search's technical requirements are eligible. Google's recommendations: allow crawling, link pages internally, keep important content as text, offer a good page experience, keep structured data consistent with visible content, and keep Business Profile information current.
• ChatGPT: OpenAI says it uses a crawler called OAI-SearchBot to surface sites in ChatGPT search, and that sites that opt out of it will not be shown in ChatGPT search answers.
• Perplexity: PerplexityBot is used to surface and link sites in Perplexity results; Perplexity says it is not used to crawl content for AI foundation models.

Beyond that, most lists of "ranking factors for AI" are guesses or conclusions drawn from small experiments. We do not present them as fact here.

What to do on your business website

None of the following is an "AI trick." All of it is work that is already right for human visitors and search engines. AI search simply rewards it, because systems need clear, checkable information to produce answers.

1. Answer "who we are, what we do, where we work" plainly

Write your business name, what you do, which cities and regions you serve, your hours and your contact details in one clear place. Give each service its own page and say in the first paragraph what the service is, who it is for and how it works. If an AI quotes one sentence, that sentence should make sense without the rest of the page.

2. Write the same facts everywhere

Your name, address, phone number and service descriptions should not contradict each other across your site, your Google Business Profile, your social accounts and directories. Contradictions confuse search engines and people alike.

3. Add structured data, with the right expectations

Structured data (schema.org markup) is a block of code that tells machines plainly what a page is about, for example "this is a local business, with this address and these hours." Google says structured data helps it understand a page and can enable rich results. It also says explicitly that no special schema is required for AI features. So treat it as a way to give accurate, complete information, not as an AI switch. One rule: the markup must match what is visible on the page.

4. Answer frequently asked questions the way they are asked

Collect the questions customers ask you by phone, DM and form, and answer them briefly and clearly. This makes the page useful to visitors and produces self-contained passages an AI system could quote. Write FAQs to help visitors, not to get a particular look in search results.

5. Keep content current

When your scope of service, hours or address change, update the page the same day. Outdated information can circulate as wrong information in AI answers too. Showing publish and update dates on articles also builds reader trust.

6. Be mentioned by others

It is thought that AI systems look not only at your site but also at what is written about you elsewhere; the weight varies by system and no exact share is published. Real customer reviews, local news and directory listings, mentions on industry sites and partner pages strengthen your reputation either way. Fake reviews or bought mentions are risky and unethical; we do not recommend them.

7. How it relates to Google Business Profile

Google recommends keeping Business Profile information current for its AI features. For a local business the profile is a source that gives Google your address, hours, category, photos and reviews directly. We do not know how much weight it carries in AI answers, but a current, consistent profile does no harm and clearly helps local search. The key point is that your website and profile say exactly the same things.

robots.txt and AI crawlers

robots.txt is a plain text file at the root of your site that tells bots which sections they may enter. The key point: AI companies do not have a single bot. They usually have different bots for different jobs, and which one you allow decides which job you allow. According to the official documentation:

• OpenAI: OAI-SearchBot surfaces sites in ChatGPT search; GPTBot collects content that may be used to train foundation models; ChatGPT-User visits pages when a user asks ChatGPT a question or uses a custom GPT. OpenAI says blocking GPTBot does not affect search appearance, but sites that opt out of OAI-SearchBot will not be shown in ChatGPT search answers. It also says robots.txt rules may not apply to ChatGPT-User. A robots.txt change can take about 24 hours to take effect.
• Anthropic (Claude): ClaudeBot collects content that may contribute to model training; Claude-User accesses sites when a Claude user asks a question; Claude-SearchBot analyzes content to improve search result quality. Anthropic says its bots honor robots.txt directives.
• Perplexity: PerplexityBot surfaces sites in Perplexity results and respects robots.txt. Perplexity-User fetches a page when a user asks a question; Perplexity says that because a user requested the fetch, it generally ignores robots.txt rules.
• Google: Googlebot crawls for Google Search and its AI features; Google-Extended is a separate robots.txt token that controls whether content may be used to train Gemini models and for grounding. Google says Google-Extended does not affect inclusion in Google Search and is not a ranking signal. To limit what shows in AI Overviews, Google points to existing controls such as nosnippet, data-nosnippet, max-snippet and noindex.

Our suggestion for small businesses: if your goal is to win customers, do not block the search and answer bots (OAI-SearchBot, Claude-SearchBot, PerplexityBot, Googlebot); blocking them means not appearing in those systems. Whether to block training bots (GPTBot, ClaudeBot, Google-Extended) is a choice: if you do not want your content used for model training you may block them, and the documentation says this does not affect search visibility. This changes quickly, so check the companies' current bot pages before deciding (sources are listed in the META file). Also, robots.txt is a courtesy rule, not a security measure; it does not stop bots that ignore it.

An honest status report on llms.txt

llms.txt is a proposed file format placed at a site's root that summarizes the important pages for AI systems. Many SEO articles present it as a must; we think that is exaggerated. What we know:

• Google's AI features guide says you do not need "new machine readable files, AI text files, or markup" to appear in those features.
• Google's John Mueller said in June 2025 that "no AI system currently uses llms.txt." That describes the situation at that time; he also suggested it might change in the future.
• We found no published, official statement that ChatGPT, Claude or Perplexity use it when choosing answers.

Conclusion: creating an llms.txt may be cheap and harmless, but there is no evidence it helps. If your time and budget are limited, invest first in clear pages, consistent information and your Google Business Profile; leave llms.txt for last. Be cautious if someone says "we will get you into ChatGPT with llms.txt."

Why measuring is hard

In classic SEO you can see which searches you rank for and how many clicks you get. Visibility in AI search is much more limited:

• Google: Search Console has a Generative AI performance report showing impressions in AI Overviews and AI Mode. It has limits: according to Google's help page the usual performance report limits (such as the 1,000 row limit) apply, and the report may not appear without enough impressions. Search Labs experiments are not included.
• ChatGPT, Claude, Perplexity and others: We know of no official dashboard that tells you how often you were recommended. What you have is seeing bot visits in server logs and checking referrals from these tools in analytics; both give an incomplete picture, because someone reading an answer and searching for you directly later stays invisible.
• Manual testing: You can ask different tools the questions your customers would ask and see whether your brand appears. But answers vary by person, time and location; a single try is a snapshot, not a result.

That is why we do not promise figures like "you grew X percent in AI," and why we suggest asking how any such figure was measured.

Common mistakes

• Blocking crawling by accident: An old robots.txt or a security plugin may be blocking search bots too. Check your file and your site's bot access once.
• Managing all AI bots as one: A training bot is not a search bot; blocking everything also means not appearing where you want to.
• Burying key information in images: If your service list or hours exist only in a picture, they cannot be read as text. Google also recommends keeping important content in text form.
• Mass-producing AI content: Dozens of thin pages saying the same thing add value for neither users nor search engines.
• Inconsistent information: One phone number on the site, another on the profile, a third address on social media.
• Believing big promises: No agency can guarantee that an AI answer will mention you, because we do not control those answers.

No guarantees: what we promise and what we do not

We cannot guarantee any ranking or any mention in an AI answer; these systems are not ours and their rules keep changing. What we can offer: keeping your site crawlable, clear and current, making your information consistent, reporting what we can measure, and saying plainly what we cannot. If you want to talk about what this would look like for your business, see our SEO, GEO and AEO service or reach us through the quote wizard.