How do you get cited by ChatGPT and AI search? The honest answer is that nobody can guarantee it. No agency, tool or setting makes an AI assistant mention your business. What you can control is whether your website is easy for these tools to reach, understand and trust. Those are the conditions that make a citation possible.

This guide explains how AI-generated answers choose their sources and what that means for a small business website. It also covers what to do in practice and how to measure the results without guesswork. It's written for businesses in the UK, US, Canada and Australia. The core principles apply in all four markets, but which AI features are available, and what each tool reports, can vary by country and change over time.

What AI search visibility actually means

AI search visibility is how often, and how accurately, your business appears in answers produced by AI-powered search tools. That includes:

  • Google AI Overviews and AI Mode, the AI-generated summaries and conversations inside Google Search
  • ChatGPT search, when ChatGPT looks things up on the web and links to its sources
  • Microsoft Copilot and Bing AI-generated answers
  • Perplexity, Claude and other assistants that search the web before answering

Visibility can mean two different things. Your page can be cited, which means linked as a source. Or your business can be mentioned, which means named in the answer itself. Both depend on the same groundwork: the tool has to be able to find your content, understand what it says and judge it worth using.

This work is sometimes called generative engine optimisation (GEO) or AI search optimisation. The labels are new, but most of the work underneath them is good, well-established SEO.

How AI-generated answers differ from traditional search results

A traditional results page shows a ranked list of links, and the searcher chooses which to open. An AI-generated answer does some of that reading for them. It pulls information from several pages, writes a summary and links to some of the sources.

That changes a few things for a business:

  • One question can trigger many searches. Google says AI Overviews and AI Mode may use a "query fan-out" technique. That means running several related searches across subtopics to build a response. A page that answers one specific part of a question well can be used, even if it wouldn't rank for the broad question.
  • Answers vary. Google notes that AI Mode and AI Overviews may use different models and techniques, so the responses and links they show will vary. The same question asked twice, or by two people, can produce different sources.
  • Not every search gets an AI answer. Google says AI Overviews appear only when its systems judge them helpful in addition to classic results. Many searches still show a normal results page.
  • Fewer, different clicks. Some searchers get what they need from the summary and never click. Google says visitors who do click through from its AI features tend to spend more time on the site. That's Google's own observation, so judge it against your own data.

Step 1: Make sure AI tools can reach your website

Nothing else matters if your pages are blocked. This is the most common fixable problem, and it's usually accidental.

Google uses the same crawler and index for its AI features as for normal search. Its guidance on AI features says there are no additional requirements to appear in AI Overviews or AI Mode. To be shown as a supporting link, a page must be indexed and eligible to appear in Google Search with a snippet. If a page is set to noindex or nosnippet, it won't be used as a source.

Other AI search tools run their own crawlers, and each provider sets its own policies. Most now separate three kinds of access:

  • Automated search crawling: a bot that visits pages so the tool can find, index and cite them in its answers.
  • Training crawlers: a separate bot that collects content which may be used to train AI models.
  • User-initiated retrieval: a page fetched because a user asked the assistant to look something up or open a link. Some providers say robots.txt rules may not apply to these requests, because a person asked for them.

The first two can usually be allowed or blocked separately in your website's robots.txt file. Examples from the providers' own documentation:

  • OpenAI: OAI-SearchBot decides whether your site can appear in ChatGPT search answers, and OpenAI's crawler documentation recommends allowing it if you want to appear there. GPTBot, which collects content for model training, is a separate setting, so you can allow search and block training. ChatGPT-User makes user-initiated visits and isn't used to decide what appears in search.
  • Perplexity: PerplexityBot surfaces and links websites in Perplexity's results. Perplexity's documentation says it isn't used to collect training data. Its separate user-initiated fetcher, Perplexity-User, generally ignores robots.txt.
  • Anthropic: Anthropic's help article describes three bots: Claude-SearchBot for search results, ClaudeBot for training data and Claude-User for user-directed requests.
  • Microsoft: Bingbot crawls for Bing. Bing's reporting covers citations in Microsoft Copilot and Bing's AI-generated summaries, and Bing says it respects robots.txt preferences.

Two things often block these crawlers without anyone deciding to. One is a blanket rule in robots.txt. The other is a security or firewall setting, sometimes switched on by default, that treats every bot as a threat. It's worth checking both, and making the decision deliberately: which tools you want to appear in, and whether you're happy for your content to be used for training. These policies change, so check each provider's current documentation before you change your settings.

Step 2: Write content that answers real questions clearly

AI tools are looking for passages that answer a question directly and accurately. Pages that do this well tend to share a few habits:

  • The answer comes first. Answer the question in the opening sentences, then explain. A page that takes four paragraphs to get to the point is harder to use as a source.
  • Headings describe what follows. "How much does a boiler service cost?" is more useful to a reader and to a machine than "Our pricing philosophy".
  • One page has one clear purpose. A service page that covers ten unrelated services gives no clear answer about any of them.
  • Specifics instead of slogans. Who you help, where, how the process works, what's included and what isn't. Marketing language like "innovative solutions" says nothing an AI tool can repeat.

Start with the questions your customers already ask on the phone and by email. Those are the questions they're now also asking AI assistants. Building pages around them is the core of content marketing that answers customer questions.

Use clear structure and internal links

A logical site structure helps search engines and AI tools understand how your content fits together. That means service pages, supporting guides and an about page, linked to each other with descriptive link text. If a guide explains a problem, link it to the service that solves it. If a service page mentions a process, link to the page that explains it.

Step 3: Show credibility and genuine expertise

AI tools draw on the same signals of quality that search engines have used for years. Google's guidance asks site owners to focus on unique, valuable content for people, not "commodity" content that repeats what every other page says. Its guide to helpful, people-first content includes self-assessment questions worth working through.

For a small business, credibility usually comes from things only you can provide:

  • First-hand experience. What you've actually seen, done and learned, described specifically.
  • Clear authorship. Who wrote the page, who runs the business and where you're based.
  • References for factual claims. Link to official or authoritative sources when you state a fact, as this article does. It helps readers check your claims, and it shows you've done the work.
  • Accuracy and dates. Keep content up to date, and correct anything that's no longer true. Outdated advice is a reason not to cite a page.

Step 4: Use structured data properly, but don't treat it as a shortcut

Structured data (schema markup) is code that labels information on a page, such as your business name, address, opening hours, articles and FAQs. It helps search engines interpret a page and can make it eligible for certain search features.

It is not a special key to AI answers. Google states that there's no special schema.org structured data you need to add to appear in its AI features. What Google does stress is that structured data should follow its guidelines and match the content visible on the page. Accurate Organization or LocalBusiness markup, and Article markup on blog posts, are sensible foundations. Markup that describes things the page doesn't actually show is a risk, not an advantage.

Step 5: Keep your business information consistent everywhere

When an AI assistant describes a business, it often draws on more than that business's website. That can include directory listings, review sites, social profiles and articles that mention the business. If those sources disagree about your services, location or contact details, the answer may be wrong or vague.

  • Use the same business name, address, phone number and service description everywhere they appear.
  • If you serve local customers, keep your Google Business Profile complete and accurate.
  • Ask satisfied customers for honest reviews on the platforms they already use.
  • Earn genuine mentions: trade association listings, local press, partner websites and talks or guides you've contributed to.

There's no published formula showing that a certain number of mentions leads to AI citations. Treat consistent, genuine mentions as a way to make sure AI tools describe you accurately when they do mention you. Don't treat them as a lever to pull. Buying links or fake reviews breaches search engines' spam policies and can do lasting damage.

Step 6: Check the technical basics

AI tools can only use content they can load and read. The same technical SEO checks apply:

  • Important pages return a normal "200" response and aren't blocked or set to noindex by mistake.
  • Key text is in the page itself, not only inside images, PDFs or content that appears only after a click.
  • Pages load quickly and work well on mobile. Google lists page experience among its recommendations for AI search, because visitors who click through need a good experience too.
  • Your sitemap is up to date, and your site is verified in Google Search Console and Bing Webmaster Tools.

What about llms.txt?

llms.txt is a proposed file that lists a website's key pages in a simple format for AI tools. It's optional and still emerging. Google says you don't need to create new machine-readable files or AI text files to appear in its AI features. It can be a tidy summary of your site, but it's not a substitute for the steps above, and it won't make up for content that's hard to access or understand.

How to measure AI search visibility

Measurement is improving, but it's still partial. Use several sources together:

  • Google Search Console. Clicks and impressions from AI Overviews and AI Mode are included in the main Performance report, under the "Web" search type. Google has also added a Generative AI performance report. It shows how often links to your site appeared in these features, broken down by page, country and device. It reports impressions, not clicks, and may not appear for sites with very little AI visibility.
  • Bing Webmaster Tools. The AI Performance report, launched in public preview in February 2026, shows how often your pages are cited in Microsoft Copilot and Bing's AI-generated answers. It also shows the phrases those tools used to find your content. It counts citations, not clicks, and Bing describes some of the data as a sample.
  • Google Analytics 4. Since 13 May 2026, GA4 has had an AI Assistant channel in its default channel group. Visits from recognised assistants such as ChatGPT, Gemini, Claude and Copilot are given the medium "ai-assistant" and grouped under that channel. Visits from assistants that aren't on Google's list may still appear as ordinary referrals. Google's own AI Overviews and AI Mode are not part of this channel. Google's channel definitions include them in Organic Search, alongside normal Google results, so GA4 can't separate them out. Combine this with tracking enquiries in GA4 to see whether those visits turn into leads. Our analytics and reporting service can set this up for you.
  • Manual checks. Every month, ask the main AI assistants the questions your customers would ask, and note whether you appear and whether the description is accurate. Keep it consistent: same questions, same order, and saved results.
  • Ask new customers. A "How did you hear about us?" question on your enquiry form catches discovery that no analytics tool can see. For example, someone who saw your name in an AI answer and then searched for you directly.

The limits of AI visibility tools

A growing number of paid tools promise to track your "AI visibility score". Be careful how you read them. Most work by sending a fixed set of prompts to AI assistants and recording the answers. Real users ask different questions, in different words, with different context and history. As Google itself notes, answers vary. A score based on a few hundred sample prompts is an indicator, not a measurement of what your customers actually see. No tool can see inside every AI conversation, and none can tell you exactly why a source was chosen.

A practical 30-day checklist

  • Week 1: Check robots.txt and any firewall bot settings. Decide which AI search crawlers to allow. Confirm your key pages are indexed in Google Search Console, and verify your site in Bing Webmaster Tools.
  • Week 2: List the 10 questions customers ask most often. Check that each has a clear, direct answer on your site, and write or improve the pages that don't.
  • Week 3: Make your business details consistent across your website, Google Business Profile and main directories. Check that your structured data is accurate and matches each page.
  • Week 4: Set up the measurement: the AI Assistant channel in GA4, the Search Console and Bing reports, a monthly manual check and a "How did you hear about us?" question.

Then give it time. Some providers say robots.txt changes are processed within about a day, but that only means their tools can now reach your pages. Being indexed, cited or mentioned isn't guaranteed, and any measurable effect usually takes weeks or months, as our guide on how long SEO takes to work explains. AI tools change how they choose sources regularly, so review your approach every few months rather than treating it as a one-off fix.

How this fits into SEO

AI search optimisation isn't a replacement for SEO, and it isn't a separate discipline you need to buy on its own. Nearly everything in this guide also improves how your site performs in traditional search: crawler access, clear content, credibility, structured data and technical health. Google says the same, recommending the fundamentals of SEO for its AI features.

At Digital Growth Journey, we're a London-based agency working with businesses in the UK, US, Canada and Australia. AI search readiness is part of our SEO service, not a separate product. We check that AI tools can reach your site, that your content answers real questions clearly and that your business information is consistent. We'll also tell you honestly what can and can't be measured. We won't promise citations or mentions in AI answers, because nobody can. You can see how this connects with the rest of what we do on our services page.

FAQs

Can anyone guarantee my business will be cited by ChatGPT or Google's AI Overviews?

No. These tools decide their sources automatically, and their answers vary between users and over time. Anyone who guarantees AI citations is promising something they can't control. What you can do is remove the barriers and give these tools clear, accurate content to work with.

Do I need special code or markup to appear in AI search?

Not for Google. It says there are no additional technical requirements and no special structured data needed for AI Overviews or AI Mode. Other tools need their search crawlers to be allowed in robots.txt. Accurate structured data is still worth having as part of good SEO.

Should I block AI crawlers to protect my content?

It depends on what you want. Most AI companies use separate crawlers for search and for training. That lets you allow your site to appear in AI search answers while opting out of model training. Blocking the search crawlers too means those tools are much less likely to link to you.

Is AI search replacing Google?

Not for most businesses' customers, at least not yet. A June 2026 SE Ranking study of 101,574 websites using Google Analytics found AI platforms accounted for 0.32% of website traffic between January 2025 and April 2026, compared with 42.75% from organic search. It's one dataset, from an SEO software company, and results vary by site, but the gap is large. AI answers are an additional route to being found, and Google's own AI features sit inside Google Search. The sensible approach is to build foundations that work for both.

How quickly will changes make a difference?

Some providers say robots.txt changes take effect within about 24 hours, but that only restores access. It doesn't guarantee that your pages will be indexed, cited or mentioned. Content and reputation changes usually take weeks or months to show any measurable effect. Measure monthly, and judge trends over a quarter rather than week by week.

Want to know whether your website is ready for AI search, and for traditional search? Request a Free Growth Audit, and we'll show you what's standing in the way.