A traditional SEO audit will tell you where you rank on Google, which pages are not indexed, and which competitors beat you on which keywords. All useful. All incomplete. 

It will not tell you whether ChatGPT mentions your brand when someone asks it to recommend a supplier in your category. It will not tell you whether Perplexity can read your product pages. It will not tell you that a competitor gets named in nine out of ten AI answers about your industry while you get named in none. 

Those are not edge cases any more. They are a growing share of how buyers research before they open a search engine at all. 

 

What Is an AI SEO Audit? 

An AI SEO audit assesses whether AI platforms can find, read and cite your content, and whether they currently do. It sits alongside a traditional audit rather than replacing it, and it answers a different question: not “where do we rank” but “do we get mentioned”. 

 

What a Traditional SEO Audit Checks 

Crawlability and indexation. Site speed and Core Web Vitals. Keyword rankings and gaps. On-page optimisation. Backlinks and authority. Technical errors. 

Every one still matters, and the foundations that help you rank also help you get cited. Crawlability and clean site structure are prerequisites for AI visibility, not alternatives to it. 

The limitation is the assumption underneath. A traditional audit assumes the outcome you want is a ranking position that produces a click. Increasingly the outcome is a mention inside an answer that produces no click at all. 

 

AI SEO Audit vs Traditional SEO Audit: What Actually Differs 

	Traditional SEO audit	AI SEO audit
Measures	Ranking position	Citation and mention
Assumes the outcome is	A click to your site	A recommendation inside an answer
Crawler focus	Googlebot, Bingbot	GPTBot, PerplexityBot, ClaudeBot, Google-Extended
Content question	Is this page optimised?	Can this page be extracted and quoted?
Competitive view	Who outranks us	Who gets named instead of us

What an AI SEO Audit Checks 

Live testing across AI platforms 

Not theory. Actual buyer questions run through ChatGPT, Perplexity, Gemini and Google’s AI Overviews to see who gets named. This is the part that surprises people most, because the answer is frequently a competitor they do not consider a threat, or an aggregator site they have never thought about. 

 

AI crawler access 

Separate from Googlebot, and separately blockable. GPTBot, PerplexityBot, ClaudeBot and Google-Extended each need permission in your robots.txt for the platform behind them to read your site. We regularly find these blocked by accident, usually through a plugin default or a blanket rule added by a cautious developer. If they are blocked, you are invisible to those platforms regardless of how well you rank on Google. 

 

Content extractability 

AI systems pull discrete, self-contained claims out of pages. Content written as one long unbroken argument, or with the key fact buried in paragraph nine, extracts badly. Content with clear headings, direct answers and explicit qualifications extracts well. This is a structural property of the page rather than a judgment about writing quality, and it is the difference between content that ranks and content that gets quoted. 

 

Entity clarity and structured data 

Whether machines can establish unambiguously what your business is, where it operates and what it sells. Schema does most of that work, and structured data carries more weight for eCommerce visibility than most people expect precisely because it removes the guesswork. 

 

Citation and mention tracking 

Which sources the AI platforms actually cite in your category, and what those sources have that you do not. 

 

Does Schema Still Help AI Search Visibility? 

Yes, but not for the reason it is usually sold. 

Google restricted FAQ rich results to government and health sites in August 2023, then deprecated the feature entirely in May 2026, with Search Console reporting removed shortly after. Anyone selling you FAQ schema on the promise of expandable dropdowns in the search results is selling something that no longer exists. 

Schema still matters because it is how machines resolve what your business is without inferring it. That has become more valuable as AI systems do more of the interpreting, not less. Add it for entity clarity, not for snippets. 

The same logic applies to review markup. Reviews about your own business, hosted on your own site, have been classed as self-serving and ineligible for star ratings since 2019. Testimonials on your homepage are good trust content for humans, and they will not produce stars in search. 

 

What We Find Most Often in an AI SEO Audit 

AI crawlers blocked in robots.txt :  The single most common finding, and the fastest to fix. 

Competitors cited, you are not : Usually because their content answers the question directly while yours explains the background first. 

No structured data, or three conflicting versions : An SEO plugin generates one, the theme generates another, someone pasted a third into the page body. Google picks one more or less arbitrarily, and the contradictions undermine all three. 

Key facts locked in PDFs and images : Specifications, pricing, certifications and comparison tables that a machine cannot read. This is close to universal among manufacturers, where catalogue data trapped in a PDF is usually the highest-return fix available. 

Content that ranks but does not extract : Pages performing fine on Google that never get quoted, because the useful sentence sits three scrolls down. 

 

How to Measure AI Search Visibility 

The honest answer is that this is less mature than rank tracking, and anyone claiming precision is overselling. What works today is triangulating three signals.

Brand mentions and citations across AI platforms, tested on a consistent set of buyer questions so results are comparable month to month.

Referral traffic from AI sources, which understates real impact because most AI-influenced visits arrive with no AI referrer attached.

Branded search volume, which tends to rise as AI visibility improves. Someone hears your name in an AI answer, then searches for you directly. The platform gets no credit in your analytics, but the effect is real and it shows up here.

None is sufficient alone. Together they give you something defensible, which is how we run SEO and GEO as one measurement system rather than two parallel reports. 

 

Do You Need an AI SEO Audit? 

If your buyers research before they buy, probably yes. Considered purchases, B2B, anything where someone compares options before committing. 

The case is strongest when you see a pattern a traditional audit cannot explain: rankings holding steady while traffic declines, branded searches rising with no campaign behind them, or prospects arriving already informed about options you never sent them. 

The case is weakest for impulse purchases and purely local walk-in businesses, where nobody is asking an assistant for a recommendation. 

Our Growth Audit includes AI visibility as a core component rather than an add-on, alongside technical SEO, content and competitive analysis. The ongoing implementation sits under AI SEO, and there are free walkthroughs of the fundamentals in our AI SEO guides. 

 

Frequently Asked Questions 

What is an AI SEO audit?

An assessment of whether AI platforms like ChatGPT, Perplexity and Gemini can find, read and cite your content, and whether they currently do. It covers AI crawler access, content extractability, structured data and entity clarity, plus live testing of real buyer questions to see which brands get named. 

 

How is an AI SEO audit different from a traditional SEO audit?

A traditional audit measures ranking position and assumes the outcome is a click. An AI audit measures citation and mention, and assumes the outcome may be a recommendation with no click at all. The technical foundations overlap almost completely. 

 

Does AI SEO replace traditional SEO?

No. Crawlability, site speed, clean structure and good content help both. AI visibility adds a layer on top rather than replacing what sits underneath. 

 

Will FAQ schema get me rich results in Google?

Not any more. Google limited FAQ rich results to government and health sites in 2023 and deprecated the feature entirely in May 2026. Schema is still worth adding, but for entity clarity and machine readability rather than snippets. 

 

How do I check if AI crawlers can access my site?

Look in your robots.txt for GPTBot, PerplexityBot, ClaudeBot and Google-Extended. If any are disallowed, the platform behind it cannot read your content. This is the most common finding we see and usually the quickest to fix. 

 

How do you track whether ChatGPT mentions your brand?

By running a fixed set of buyer questions across the platforms on a regular schedule and recording which brands get named, then reading that alongside AI referral traffic and branded search volume. No single metric captures it, so the three are read together.