AI Visibility Audit: Find Out Whether AI Assistants Can See Your Business

An AI visibility audit checks whether the assistants people now ask for recommendations, ChatGPT, Microsoft Copilot, Google Gemini and Perplexity, can reach your website, read your answers and cite you as a source. The audit works through six layers: AI crawler access in your robots.txt file, whether your answers sit in plain extractable HTML, your AI citation footprint across the four assistants, your Bing Webmaster Tools AI Performance data as a measurable baseline, brand mentions and entity consistency, and your structured data.

The report sets out what was found on each layer, ranks the findings by how much visibility they cost you, and gives a fix list in the order the work should be done. A full AI visibility audit costs £800 to £1,200 and is delivered in 10 to 14 business days, with a shorter AI visibility check available at £300 to £600.

What Is an AI Visibility Audit?

An AI visibility audit is a diagnostic. It answers one question in detail: when someone asks an AI assistant for a business like yours, what stops your name from coming back? That question has a technical half and an editorial half, and the audit treats both. The technical half decides whether an assistant can get to your pages and parse them. The editorial half decides whether, having read them, the assistant has anything quotable to work with.

The reason an AI visibility audit exists as a separate piece of work is that the failure is silent. A blocked crawler does not send a warning. A page that renders its content with JavaScript still looks perfect to you in a browser. Your Google rankings can hold steady while you vanish from the answers a growing share of your customers see first. If you have ever wondered whether your business is visible in AI at all, the honest answer is that you cannot know without checking, and checking is what this audit does. It applies the same evidence-first approach as the rest of our AI SEO work, aimed squarely at the retrieval route that decides who gets cited.

What Does an AI Visibility Audit Check?

Six layers, each checked separately, because a site can pass five of them and still be invisible. Every layer in the AI visibility audit produces its own findings, and each finding names the page, file or profile it came from.

Layer One: Can AI Crawlers Reach Your Site at All?

The audit reads your robots.txt file line by line and records the rule applied to each named AI user agent: GPTBot and OAI-SearchBot from OpenAI, ClaudeBot and Claude-SearchBot from Anthropic, PerplexityBot from Perplexity, and the Google-Extended token from Google. It also checks for blanket disallow rules and for CDN or firewall rules that block these agents even when robots.txt permits them, which is a common and invisible conflict. Spelling matters here, because a rule aimed at GPT-Bot rather than GPTBot silently does nothing at all.

The distinction between training crawlers and search crawlers is the part most sites get wrong, and it is the first thing an AI visibility check should establish. Blocking a training crawler keeps your content out of a model’s general knowledge. Blocking a search crawler removes you from the live AI citations an assistant produces when it goes looking for an answer, which is the visibility that actually sends you enquiries. The two are separate decisions, controlled by separate lines, and OpenAI documents its own crawlers publicly so the choice can be made deliberately rather than inherited from a 2023 template.

Layer Two: Do Your Answers Exist in Plain, Extractable HTML?

The AI visibility audit fetches your key pages the way a crawler does, without running JavaScript, and compares what comes back against what you see in a browser. Anything that only appears after a script runs, or that sits hidden inside a tab, accordion or modal, is treated as absent. This single check accounts for more invisible businesses than any other, particularly on sites built with modern front-end frameworks or heavily plugin-dependent page builders.

The second part of this layer is whether an extractable answer exists at all. A page that opens with brand positioning and reaches the actual answer in paragraph nine gives an assistant nothing convenient to quote. Pages that state the answer plainly near the top, in prose, are the ones that earn an AI citation.

Layer Three: What Is Your AI Citation Footprint Across the Four Assistants?

This is the layer most owners have in mind when they ask us to check AI citations for their business. The audit runs a fixed set of prompts, built from how your customers actually phrase their needs, across ChatGPT, Copilot, Gemini and Perplexity, and records what comes back. We log whether you appear, whether you are cited with a link, which page was cited, and which other businesses appear alongside you. The prompt set is kept so it can be rerun later against the same wording.

This part of the AI visibility audit is sampling, not measurement, and the report says so plainly. Assistant answers vary by session, by account and by location, so a single run proves very little. A consistent pattern across a full prompt set is the signal worth acting on.

Layer Four: What Does Bing Webmaster Tools Say About Your AI Citations?

Bing Webmaster Tools includes an AI Performance report showing how often your content is cited in AI generated answers across Microsoft Copilot, Bing AI summaries and partner integrations, along with the grounding queries behind those citations. It is free, first-party and repeatable, which makes it the closest thing to a real baseline currently available and the only place you can check AI citations against a published number rather than an impression. Microsoft documents the report and its metrics in its own help centre.

Its limits are stated in the audit rather than glossed over. The report covers Microsoft surfaces, not Gemini or Perplexity, and it counts AI citations rather than clicks. The audit verifies your property, establishes the starting figures and hands you the dashboard, so progress after the fix work is something you can see yourself.

Layer Five: Are Your Brand Mentions and Entity Details Consistent?

Assistants build a picture of your business from every mention of it, not only from your website. This layer of the AI visibility audit checks how your business name, address, service description and founding details appear across directories, professional bodies, press coverage, review platforms and your own social profiles, and flags every version that contradicts the others. Two spellings of your trading name or three different addresses give a model good reason to treat your identity as uncertain and cite someone clearer instead.

Layer Six: Is Your Structured Data Doing Any Work?

The audit validates the schema markup on your key pages and checks that it agrees with the visible content. Organisation, LocalBusiness, Service, Article and FAQPage markup are checked against the schema.org vocabulary, along with sameAs links tying your profiles together. Markup is not a shortcut to an AI citation, and the report will not pretend otherwise, but it removes ambiguity about who you are, where you operate and what you sell.

A similar discipline runs through our technical website audit and our standard search engine optimisation audit. The difference is the reader. Those audits ask how a search engine ranks your pages. An AI visibility audit asks whether a language model can use them.

What Does the AI Visibility Audit Report Contain?

Three parts, in this order.

Findings. Every layer written up with what was checked, what was found and the evidence behind it: the exact robots.txt lines, screenshots of rendered against raw HTML, the prompt log with the answers you were and were not in, your Bing AI citation figures, the entity inconsistencies, and the schema validation output.

Priorities. The findings ranked by cost to your visibility rather than by ease of fixing. A blocked search crawler sits above a missing FAQ schema every time, because one removes you entirely and the other makes you slightly harder to parse.

The fix list. Specific, assigned and sequenced. Each item says what to change, which page or file it lives in, whether it is a developer task or a content task, and what you should expect to see afterwards. Items we can do for you are marked as such, and items your own team can handle in an afternoon are marked too, because you should not pay an agency for work you can finish yourself.

Every AI visibility audit also includes a walkthrough call, where somebody talks you through the findings and agrees what happens in the first thirty days. A report delivered by email and left to decode alone is a file, not a diagnosis.

What Does an AI Visibility Audit Cost and How Long Does It Take?

Three tiers, priced on the same basis as our other audit work: how many pages carry your commercial intent, how many assistants and prompts the citation sampling covers, whether your site renders content client side and therefore needs deeper technical checking, and whether your Bing Webmaster Tools property is already verified or has to be set up.

A short AI visibility check suits a small site with clean HTML where you mainly want to know whether anything is blocking you. The full AI visibility audit is the right choice where you intend to act on the findings, because the citation sampling and the entity work are what turn a list of problems into a sequence of jobs. Where the Bing baseline has to be established from scratch, the AI citation figures in your report will be thin, and we will tell you that rather than present an empty dashboard as a finding.

What Do Businesses Usually Discover in an AI Visibility Audit?

Three findings come up again and again, and none of them are the ones owners expect.

The first is content that is invisible to crawlers. Service descriptions, pricing tables and FAQ answers built inside tabs, accordions or JavaScript-rendered components look complete on screen and return nothing to a crawler that does not run scripts. The business believes it has answered the question. The assistant has never seen the answer.

The second is a blocked crawler nobody remembers blocking. Disallow rules added years ago, often by a previous developer acting on advice about protecting content from AI training, are still sitting in robots.txt removing the site from live AI citations as well. Sometimes the block is at the CDN or firewall layer, where robots.txt looks entirely correct.

The third is the absence of an extractable answer. The pages rank, the crawlers get through, and there is still no clean two or three sentence statement of what the business does, where it does it and who it is for. An assistant asked to recommend somebody needs a sentence it can lift. Where there is nothing to lift, it lifts a competitor’s.

What Happens After the AI Visibility Audit?

The report is usable on its own, and a number of clients take it to their own developer and content team. Where you would rather the work was done, there are three paths. Technical remediation covers the crawler rules, rendering changes and schema implementation. Content remediation rewrites the key pages so the answers sit in plain HTML near the top, with the question phrasing people actually use. Ongoing monitoring reruns the prompt set and reads the Bing report on a set cycle, so the change in AI citations is measured rather than assumed.

Businesses working with us on AI SEO in Northern Ireland usually start with an AI visibility audit and move into remediation on the same fix list, which keeps the priorities intact. Where the wider question is how AI fits into your marketing rather than just your visibility in it, that sits with our work on AI in marketing operations.

The episode above covers how client expectations around AI have shifted and where the practical limits sit, which is useful context before commissioning any AI visibility work.

AI Visibility Audit FAQs

Ready to Check Where You Stand?

Tell us your website address and the questions you would expect a customer to ask an assistant before they get in touch, and we will come back with the right AI visibility audit tier for your site and a fixed price within the ranges above.