How to Run AI Visibility Audits & Sell Them as a Service

Your client decided to ask ChatGPT who the best company in their industry is, saw three competitors in the answer, and arrived in your inbox a few minutes later with: “Why aren’t we showing up here?”

Along with a natural onslaught of follow-up questions: Can we rank in ChatGPT? Why is Google citing them instead of us? Is this part of SEO? Are we meant to be doing something about it? Can you fix it?

And there, dear agency owner, is a pretty obvious service opportunity. An AI visibility audit gives you a way to answer those questions (and get paid to do it). 

But what you may not have yet is the process.

That’s what we’re sorting here. I’ll go through what to check in an AI visibility audit, ways to package and sell it, and how it can lead into services you already offer.

What is an AI visibility audit, then?

An AI visibility audit checks whether, where, and how a brand appears when people use AI-powered search tools to research its services, products, industry, competitors, or buying questions.

Depending on the client, you might check Google AI Overviews, Google AI Mode, Claude, ChatGPT, Gemini, Perplexity, or whichever platforms their customers are likely to be using.

You’re looking for things like brand mentions, citations, recommended companies, pages linked, competitor appearances, and the questions where your client is nowhere to be seen.

Tools are getting far better at helping with this. Ahrefs’ Brand Radar tracks mentions, citations, impressions, AI share of voice, cited pages, competitors, and custom prompts across several AI platforms.

Semrush’s AI Visibility Toolkit does a similar job, with brand visibility, competitor analysis, prompts, mentions, citations, and AI search reporting.

Basically, an SEO audit asks: Are we being found in search?

An AI visibility audit adds: Are we being mentioned, linked to, or recommended in AI-generated answers?

And, when we aren’t: Who is?

That last question tends to get clients’ attention.

Funny that.

Should your agency offer AI visibility audits?

If AI search already overlaps with what you sell, I’d say yes. Especially since most of the work that comes out of an AI visibility audit won’t be new to you.

Google said AI Overviews and AI Mode still run off the same Search ranking and quality systems. Translation: keep doing the SEO properly and publish original, helpful content.

So this isn’t a case of throwing everything you know about SEO in the bin because AEO (Answer Engine Optimization) is now a thing.

Good news all round.

And if you want the full rundown on what Google is looking for before a page gets cited, have a read of my other guide on how to rank in AI Overviews.

For now, let’s talk AI visibility audits: what’s involved and how to sell it.

#1 Ask AI tools the questions a real customer might ask

Start by taking the kinds of questions a potential customer would ask and putting them into the AI tools you’re auditing (ChatGPT, Google AI Mode, Gemini, Perplexity, and so on).

For an accountancy firm, you might check questions around finding an accountant, choosing between providers, pricing, software, specialist sectors, or common problems clients need help solving.

You can pull those questions from Google Search Console, keyword research, sales calls, support messages, customer conversations, and the FAQs the team answers all the time.

When you’re typing in questions, think to yourself: would a potential buyer genuinely ask this?

Someone probably won’t type “Jones Accounting Ltd.” 

It’ll look more like:

• “Who are the best accountants for small creative agencies in LA?”
• “Are Jones Accounting in LA any good?”
• “Based on what you know about my company, what local accountant should I use?”

#2 Benchmark their AI visibility score against competitors

Once you know what you’re checking, get a baseline.

• Where does the client appear?
• Where don’t they?
• Are they showing up for branded questions but disappearing when someone asks for a provider?
• Are they getting mentioned by name but never linked as a source?
• Is nearly all their AI visibility coming from just one blog post?
• Are the same few competitors showing up everywhere?

Google made this easier in June 2026 when it launched a dedicated Generative AI performance report in Search Console. For sites with access (some are still waiting patiently), the report shows impressions from AI Overviews and AI Mode, the pages receiving those impressions, plus country, device, and date data.

Now, you don’t need to overcomplicate the findings.

A quick snapshot can be enough:

• Strong branded visibility.
• Weak for category searches.
• Missing from three high-intent questions.
• Competitor A keeps showing up instead.

There.

Now everyone knows what problem you’re talking about.

#3 See who keeps getting picked up instead

Look at who shows up when the client doesn’t.

That could be direct competitors, larger brands, directories, publishers, forums, review sites, or industry bodies.

Then start figuring out why they’re all popping up instead. 

Maybe one competitor has much stronger comparison pages. Another publishes pricing while your client keeps theirs under lock and key.

You might find their people quoted across industry websites, hundreds of detailed reviews, or compelling case studies packed with specifics.

Meanwhile, your client has zero reviews and a case-study section made up of three logos.

If you want to speed things up, Ahrefs Brand Radar can benchmark AI share of voice against competitors and surface the pages and domains getting cited. Very handy!

#4 Look at what’s getting cited in AI recommendations and Google AI Overviews

Now inspect the sources behind the answers you’re getting across AI engines. What kind of pages keep coming up?

• A detailed guide?
• A service page?
• A pricing page?
• Original research?
• A comparison?
• A case study?
• A directory?
• A Reddit thread?

Then compare those sources with the content your client has already published.

Say your client has a service page called “Recruitment Services for Employers.” It’s got lots of “people-first” language, but very little detail.

Their competitor has published “How Much Does a Recruitment Agency Charge?” And on their page, they’ve talked about fee models, percentages, retained versus contingency, examples, and when each option suits.

One answers the buying question. One…doesn’t.

Guess which one keeps getting picked by AI platforms?

#5 Go through the content they already have

Now for the fun bit. You read the website content.

I know. Not actually that fun (sorry).

You’ll want to check the pages tied to the questions you’ve been tracking.

• Do they answer something specific?
• Can someone find the answer quickly?
• Do they cover the obvious follow-up questions?
• Is there first-hand information?
• Are claims backed up?
• Are sources linked?
• Is the author named?
• Has anyone updated the article in the last few years?
• Are related pages linked together?

And, probably the biggest question:

• Is there anything on this page that couldn’t have come from another company in the industry?

Google is very clear on this point. Its guidance for AI search pushes original, “non-commodity” content over another rehash of what everyone else has already said.

So if the client has 60 blogs that all sound like: “Running a business in a fast-moving world can be challenging. Here are five benefits of working with a professional…”

Congratulations.

You’ve found some more work to pitch and get paid for.

#6 Make sure the site is ready for Google AI Mode

There’s no point building a 40-prompt tracking sheet while Google still can’t index the page you actually want cited. Priorities. 

So make sure Google can access the client’s strongest pages:

• Can it crawl the page?
• Can it index it?
• Are the canonicals right?
• Can it reach the page through internal links?
• Does the main content appear as readable text?

That’s straight from Google’s generative AI search guidance.

#7 See where else your client’s brand shows up

Your client’s website is obviously going to say nice things about them. Now you need to find out whether anyone else is saying the same thing.

Check reviews, social media platforms, directories, trade publications, podcasts, partner websites, guest articles, YouTube, press coverage, forums, and the client’s Google Business Profile.

You want to make sure the places they appear tell a clear, consistent story about who they are and what they’re good at.

If the website positions them as a non-profit specialist, but every other profile says “full-service digital agency,” AI tools don’t know which version of the business to run with. 

Flag outdated bios, directory listings, profiles, and any other old wording that’s still hanging around.

You should also look at where competitors are getting mentioned and your client isn’t. If the same names keep showing up in industry articles, roundups, podcasts, or reviews, that’s worth noting for future PR work.

#8 Give your client their report, what it means, and the fixes needed

Start with where they currently stand.

Then show which questions and platforms you checked, where they show up, where competitors beat them, which pages or content gaps seem to be contributing, and what you recommend doing first.

Separate the recommendations into do this now, do this next, and worth looking at later.

Then explain how you’ll measure any changes.

#9 Let the AI visibility audit lead to more work

One of the best things about this kind of audit is how naturally it feeds into services you probably already sell.

• Bare-bones website copy can become a rewrite project.
• Missing buyer questions can turn into comparison pages, pricing guides, or FAQs.
• Not enough proof can lead to new case studies.
• Technical problems can become SEO improvement plans.
• Inconsistent brand information across the web can become a directory and profile clean-up.
• Limited third-party mentions can turn into guest blogging or digital PR.

The audit gives you the why. Your existing services give you the what next.

P.S. If clients are starting to ask about AI visibility but you don’t fancy building the audit process, reviewing all the content, and writing the report yourself, you can always white-label the job with… shameless plug incoming…  The Content Lab.

You keep the client relationship, get a new service to offer, and still make a profit. Nice!

How do you actually sell an AI search visibility audit?

Please don’t open the sales conversation with “Hello madam, would you like to buy an AEO audit?”

Unless your client spends their free time listening to SEO podcasts for fun, there’s a good chance they have absolutely no idea what that means.

Sell the problem they already recognise instead:

“Your competitors are showing up in AI answers and you’re not. We can find out where you’re being missed, what they’re doing differently, and what you can do next.”

Much more likely to land with them.

Package it so they know what they’re buying

I’d start with a fixed-scope, one-off audit rather than another monthly retainer bill they need to justify.

Spell out exactly what’s included: which AI platforms you’ll check, how many buyer questions you’ll review, whether competitor research is included, and what they’ll get at the end.

You could even offer two levels:

• AI visibility snapshot: A smaller check for existing clients who want to know where they stand.
• Full AI visibility audit: Deeper prompt research, competitor comparisons, source analysis, website review, and recommendations.

Clients are far happier buying something when they can picture what turns up at the end.

Sell it when the problem is already in front of you

Your easiest prospects are probably already on your books. Bring it up when a client asks why a competitor appears in ChatGPT, during an SEO review, before a website rebuild, or when you’re already auditing their content.

Expect a few objections

And get ready to answer them:

“Can’t we just check ChatGPT ourselves?”

Of course. But checking a few prompts yourself only tells you what showed up in that moment. An audit looks across the questions buyers are asking, compares competitors, checks the sources being cited, and works out what’s probably helping or holding your brand back. 

“Can you guarantee we’ll show up?”

No. And I’d be wary of anyone who says they can. What we can do is improve the things AI search systems are looking at: the content, technical setup, authority signals, proof, and how well your brand is understood across the web. 

“Do we really need this yet?”

If your customers are using AI search and competitors are already appearing there, waiting until everyone else has a head start probably isn’t the best plan. 

“Isn’t this just SEO with a new name?”

There’s plenty of crossover, which is good. The difference is that we’re also checking whether your brand is being mentioned, linked to, and actually recommended inside AI-generated answers, and which competitors or sources keep appearing instead. 

Most of this is work you already know

If you already do SEO, content, or website strategy, you’re not starting from scratch here. Take what you already know, add the AI visibility layer, and start testing it with clients.

And if you want somewhere to compare notes with other agency owners figuring all of this out, come hang out with us in The Admin Bar community. You’ll find web designers, developers, and marketers swapping ideas, asking questions, and sharing what they’re seeing.

For deeper support, extra resources, and more of the “oh thank God, someone else is dealing with this too” kind of stuff, Barflies is where the extra goodies live.