How to Get Your Business Cited by AI Assistants Like ChatGPT and Gemini

AI assistants cite websites that are easy to read, easy to verify and easy to attribute. That means clear direct answers near the top of a page, structured data that matches what the page actually says, a site that AI crawlers are allowed to fetch, and a business that is described consistently wherever it appears online. Keyword density and backlink counts matter far less here than they do in classic search.

If people are starting their research in ChatGPT, Gemini, Perplexity or Google's AI Overviews, being invisible there is a commercial problem, not a technical curiosity. Here is what actually moves the needle.

First, check you are not blocking them

This is the step most businesses skip, and it makes everything else pointless. AI crawlers are separate from Googlebot and are frequently blocked by default — particularly by CDN and security layers such as Cloudflare, which added bot controls that many site owners never reviewed.

Check two places:

1. Your robots.txt — look for rules targeting GPTBot, ClaudeBot, PerplexityBot, Google-Extended or CCBot.
2. Your CDN or firewall — in Cloudflare this sits under AI Crawl Control or Bot Management. A site can be fully open in robots.txt and still blocked here.

Decide deliberately. Some publishers block AI crawlers to protect content, which is a legitimate choice. What you should not do is block them by accident and then wonder why you are never mentioned.

Answer the question in the first paragraph

Language models extract passages. A page that opens with three paragraphs of throat-clearing before reaching the point gives them nothing clean to lift.

Write the way a good reference book does: state the answer, then explain it. If the page is called "How much does X cost", the first sentence should contain a number. If it is "What is Y", the first sentence should define Y. This also happens to be better for human readers, which is the usual sign that an SEO tactic is sound.

Make your structured data tell the truth

Schema markup is how a machine confirms what your page claims. The types that matter most for a service business:

• Organization — who you are, where you operate, how to reach you.
• Service — what each service page actually offers.
• FAQPage — question and answer pairs that mirror the visible text.
• Product — for anything with a price.

One rule governs all of them: the markup must match what a visitor can see on the page. Schema describing prices, ratings or services that do not appear in the visible content is a liability, not an advantage.

Be described the same way everywhere

Models build a picture of your business from many sources, not just your website. If your company is "IM Holding UK" on your site, "IM Holding Ltd" on LinkedIn and "IM Digital" in a directory listing, you have made yourself three half-entities instead of one confident one.

Audit the basics — legal name, trading name, address, phone number, description of what you do — across your website, LinkedIn, Companies House, industry directories and any press coverage. Consistency here does more for AI visibility than most on-page work.

Earn mentions, not just links

Classic SEO taught everyone to chase links. Generative engines lean more heavily on being mentioned by sources they already trust — trade publications, industry roundups, genuine reviews, podcast appearances, conference listings.

A paragraph about your company in a respected trade title is worth more here than a batch of directory links, because it gives a model corroborated, attributable text about you.

Publish an llms.txt file

An llms.txt file at your domain root points AI systems to your most important pages, in the same spirit as a sitemap. The standard is young and not every model honours it yet, but it costs very little to add and there is no downside.

Several SEO plugins now generate one automatically. If you already have a file that was created by something else, check its contents before trusting it — auto-generated versions sometimes include raw page-builder code instead of readable text, which defeats the purpose entirely.

Measure it, or you are guessing

AI visibility is not yet reported in Search Console the way clicks are. To track it, do two things:

• Ask the questions yourself. Build a list of 20–30 questions a customer would ask, run them across ChatGPT, Gemini, Perplexity and Google AI Overviews on a fixed schedule, and record who gets cited. It is manual, and it is the most honest signal available.
• Watch referral traffic. Visits from chatgpt.com, perplexity.ai and similar are small today but they are real, and they are growing.

What not to bother with

Ignore anyone selling guaranteed AI citations. No one controls what a model says, the systems change frequently, and claims of guaranteed placement are either misunderstandings or marketing. The durable work is the same work that has always paid: be genuinely useful, be clear, be verifiable, be consistent.

A short checklist

1. Confirm AI crawlers are not blocked in robots.txt or at your CDN.
2. Rewrite the opening paragraph of your most important pages to answer the question directly.
3. Add Organization and Service schema, and FAQPage where you have real questions.
4. Make your business details identical across every platform you appear on.
5. Pursue mentions in publications your customers already read.
6. Add an llms.txt file and check what is in it.
7. Set a monthly reminder to run your question list and record the results.

If you would like us to run that audit on your site, get in touch — or read our guide to what SEO costs in the UK if you are weighing up budget.