An AI visibility audit checks whether AI assistants understand your business well enough to name, cite or recommend it when a potential customer asks a relevant question.
That means testing more than whether your website contains the right keywords.
A proper audit looks at:
- the questions your buyers actually ask
- whether your business appears in the answers
- which competitors appear when you do not
- whether AI systems cite your website or third-party sources
- whether your business information is consistent
- whether your pages answer questions clearly enough to be extracted
- whether technical issues make your content difficult to access
At Lobo Digital, the manual audit goes further than an automated score. Real buyer questions are run across multiple AI assistants, then the answers are reviewed for mentions, citations, competitor visibility and the signals that may explain the result. The standard GEO coverage includes ChatGPT, Perplexity, Gemini and Claude.
The important word is measure.
An audit can show where you stand today and what appears to be holding you back.
It cannot promise that ChatGPT, Gemini, Claude or any other system will recommend you tomorrow.
Table of contents
- What is an AI visibility audit?
- Why the audit starts with buyer prompts
- Which AI assistants are tested?
- What do the answers actually reveal?
- What does the audit check on your website?
- What are entity and citation checks?
- What does the report contain?
- How is the free checker different from the full audit?
- What an AI visibility audit cannot promise
- What happens after the audit?
- Frequently asked questions
What is an AI visibility audit?
An AI visibility audit is a structured review of how your business appears across AI-powered search and recommendation systems.
Traditional SEO audits usually ask questions such as:
Can Google crawl this page?
Which keywords does it rank for?
Are the title tags and internal links correct?
Those questions still matter.
An AI search audit adds another layer:
If a buyer asks an AI assistant who they should use, does your business appear in the answer at all?
That distinction matters because AI assistants do not always present ten ranked links and ask the user to choose.
They may provide a short answer containing only a handful of businesses or sources.
OpenAI itself says ChatGPT Search uses multiple factors to surface relevant and reliable information and explicitly states that placement is not guaranteed.
OpenAI’s official guidance on ChatGPT Search visibility
That is why an audit should measure actual visibility rather than assuming a strong Google ranking automatically transfers into AI recommendations.
Why the audit starts with buyer prompts
The first mistake in AI visibility testing is asking the assistant the question you want people to ask rather than the question buyers actually use.
A buyer looking for a commercial cleaning provider is unlikely to type:
Tell me which business has the strongest generative engine optimisation signals in commercial cleaning.
They are more likely to ask:
Who are the best office cleaning companies in London?
or:
Which cleaning company can handle a 50,000-square-foot office?
The phrasing changes the context.
A useful audit therefore starts by building a named prompt set around:
- your service
- your sector
- your location
- common buying requirements
- comparison questions
- high-intent problems
- the language customers actually use
The prompt set creates a repeatable benchmark.
Instead of asking vaguely whether “ChatGPT knows us”, you can track a defined group of commercial questions over time.
That approach also connects directly with the wider Ranked Not Recommended problem: ranking well traditionally does not automatically mean being included when an AI assistant creates a shortlist.
Read the Ranked Not Recommended campaign
Which AI assistants are tested?
Lobo Digital’s standard GEO work covers four assistants:
The same prompt is useful across all four because the point is not to find one flattering answer.
It is to see whether a pattern exists.
One assistant may name you.
Another may name three competitors.
Another may know the company but describe its services incorrectly.
Another may cite an old directory instead of your website.
Those differences are information.
They show whether the problem appears isolated to one system or whether there is a wider issue with how the business is represented online.
Different assistants also use different search, retrieval and answer-generation systems. Claude, for example, can search the live web and provide source citations when web search is used.
Anthropic’s official explanation of Claude web search
That is one reason testing a single assistant is too narrow for a useful audit
What do the answers actually reveal?
An audit should record more than whether the company appears.
For each buyer prompt, look at several separate signals
Are you named?
The simplest question.
Does the assistant actually include your business?
Are you cited?
Being named and being used as a source are different.
An assistant may mention your company while citing someone else.
It may also cite your content without recommending the company itself.
Both observations matter.
Who appears instead?
This is often the most commercially useful part of the exercise.
If the same two competitors repeatedly appear when your business does not, there is something to investigate.
Look at:
- What their websites say more clearly
- Where they are mentioned externally
- Which sources support their claims
- How consistent their business details are
- Whether their service pages answer the prompt more directly
The audit is therefore not just a report about your website.
It is also a competitor visibility benchmark.
Is the answer accurate?
Being mentioned is not enough if the assistant says the wrong thing.
Check whether it understands:
- Your services
- Sectors
- Locations
- Specialisms
- Company name
- Positioning
Incorrect descriptions often point towards an entity-consistency problem rather than a simple content problem.
What does the audit check on your website?
The answer from the assistant is the visible symptom.
The next job is finding possible causes.
A proper audit should therefore review the website itself.
Content and answer formatting
Can a model find a direct answer to the buyer’s question?
Or does the page spend 300 words introducing the company before explaining what it actually does?
Important answers should be clear, specific and easy to extract.
Structured data
Does the site explicitly describe the organisation, services and relevant business information in machine-readable markup?
The Lobo Digital checker reviews structured data as one of its six core areas.
Technical accessibility
Can relevant crawlers reach the content?
Is important text present in accessible HTML?
Are robots.txt, JavaScript, CDN or firewall rules creating unnecessary barriers?
Content depth
Does the website answer the questions customers actually ask, or does it simply describe services at a high level?
These are different things.
A service page can be perfectly polished and still give an AI system very little useful material for answering a specific buyer question.
What are entity and citation checks?
An AI visibility audit also looks beyond your website.
That matters because your company is not defined online only by what it says about itself.
Entity clarity
The audit asks whether your business looks like one consistent entity across the web.
That includes signals such as:
- Company name
- Services
- Locations
- Founder or leadership details
- Business profiles
- Sector associations
If those details contradict each other, an AI system has more ambiguity to resolve.
Third-party corroboration
The audit also looks for sources outside your own domain.
For example:
Google’s own guidance for its AI Search features continues to emphasise the same fundamental requirement it uses elsewhere: useful, reliable content that can be discovered and supported through normal Search systems. Google also states that there are no special additional requirements that guarantee inclusion in AI Overviews or AI Mode.
Google’s official guidance for AI features in Search
What does the report contain?
The full audit is designed to answer a practical question:
Where are we being left out, and what should we fix first?
Lobo Digital describes its full manual audit as a senior person running your real buyer questions across the engines and reviewing what comes back. The current service page lists a turnaround of around three working days, with a written report covering the questions tested, the engines checked and competitors named instead of you. The audit is currently offered free, and the report remains yours whether or not you go on to work with Lobo Digital.
The report should give you:
- the prompt set tested
- visibility by assistant
- where you are named
- where you are cited
- where competitors appear instead
- inaccurate or inconsistent brand information
- website-level issues
- entity or external-authority gaps
- priority fixes
The point is not another score with no explanation.
It is a list of observable problems that can be acted on.
How is the free checker different from the full audit?
These are related, but they are not the same product.
AI Visibility Checker
The checker is automated.
You provide:
- your website
- your sector
- up to three questions a buyer might ask
It takes around 30 seconds and gives an immediate visibility score plus three findings. The fuller checker report includes more detailed findings, a priority fix and quick wins. There is no signup required to see the initial score.
Full AI Visibility Audit
The audit is manual.
A person reviews your real buyer questions across the assistants and interprets the answers rather than relying only on an automated score.
Use the checker when you want:
Is there likely to be a problem here?
Use the audit when you want:
Where exactly is the problem, who is appearing instead of us, and what should we do about it?
If you are comparing software platforms and tracking tools, link to the dedicated AI visibility tools comparison rather than turning this service explainer into a software review.
What an AI visibility audit deliberately does not promise
This section matters.
An AI visibility audit cannot guarantee:
- a ChatGPT recommendation
- a Gemini mention
- a Claude citation
- a fixed AI “ranking”
- identical answers for every user
- a permanent position
- a specific lead volume
Generative answers can vary.
The systems change.
The underlying web evidence changes.
And no agency has access to a private switch that forces an assistant to recommend a company.
OpenAI explicitly says placement in ChatGPT Search is not guaranteed, even when a website is eligible for inclusion.
A credible audit should therefore give you:
evidence, diagnosis and priorities.
Not a promise it cannot control.
This is also why Lobo Digital’s GEO service states plainly that it does not guarantee ChatGPT recommendations.
What happens after the report?
That depends on what the audit finds.
You may need:
- stronger service-page answers
- clearer entity information
- structured data
- technical crawler fixes
- third-party citations
- stronger supporting content
- a better internal linking structure
- traditional SEO work before GEO work
Sometimes the result is:
You are already in good shape.
Sometimes it is:
The site is technically fine, but competitors have far stronger independent corroboration.
And sometimes:
There is no point starting an authority campaign until the core website is fixed.
The report should make that distinction clear.
For the wider service approach:
Explore Lobo Digital’s GEO and AEO services
Or for the deeper manual review:
Read about the full AI Visibility Audit
Frequently Asked Questions About AI Visibility Audits
Common questions about how AI visibility audits work, what they include, how long they take and what happens after the report.
How long does an AI visibility audit take?
Lobo Digital currently states that the full manual audit takes around three working days.
The automated AI Visibility Checker is much faster and returns the initial score in roughly 30 seconds.
Is the AI visibility audit really free?
Yes.
Lobo Digital currently offers both the instant checker and the deeper manual visibility audit free.
The checker does not require signup to see the initial score.
What do you need from me?
For the instant checker, you provide your website, sector and up to three questions a potential customer might ask AI when looking for a business like yours.
For the manual audit, the prompt set can go deeper into services, locations and commercial buying questions.
What happens after the report?
You keep the report whether or not you work with Lobo Digital.
If problems are identified, the next step depends on the evidence. It may involve GEO, AEO, technical fixes, content changes, entity work or traditional SEO foundations.
Does the audit cover Bing and Copilot?
The standard four-assistant GEO set described by Lobo Digital is ChatGPT, Perplexity, Gemini and Claude.
Lobo Digital states that other engines can be added when they are relevant to the buyer audience. Google AI Overviews is treated separately through its AEO work.
If Copilot is important to how your customers research suppliers, include it in the audit brief rather than assuming the standard four cover every platform.
Can an AI visibility audit guarantee that ChatGPT will recommend my business?
No.
No legitimate audit can guarantee a future AI citation or recommendation.
The audit can identify whether you are currently being named, what competitors appear instead and which observable technical, content or authority gaps deserve attention.
It cannot control the answer an independent AI system generates.
Know what AI sees before you optimise for it
An AI visibility audit should not be mysterious.
It is a structured way of answering:
What are buyers asking?
Which businesses are the assistants naming?
Where do we appear?
What sources are being used?
What signals may explain the gap?
From there, you can decide whether the problem is technical, structural, entity-based, authority-based or simply that your competitors currently provide stronger evidence.
That is much more useful than buying a GEO programme because somebody told you that AI search is the future.
Measure first.
Then decide what deserves fixing.
If you want the quickest starting point, run Lobo Digital’s free checker and see what comes back.
Run the free AI visibility check
