Last updated: September 2026 by Md Afraz Alam, Founder of Afraz Alam Marketing Services.
An AI visibility audit (the full name is AI search visibility audit) is the fastest way to find out if ChatGPT, Google's AI Overviews, and Perplexity actually know your business exists. Type your business name into ChatGPT right now and ask it what you do. Then ask it who the best option is for the exact service you sell, in your exact city or niche. Whatever comes back, or does not come back, is closer to the truth about your visibility than your Google ranking is.
Get a free AI visibility check
I run the 5-question audit below on your business live, on a free 30-minute call, and tell you which of the 3 engines is the problem. US businesses only. No pitch deck.
Most business owners have never done this. They check their Google ranking, maybe glance at Search Console, and assume that if the website looks fine there, it is fine everywhere. It is not the same test anymore. Google's own AI Overviews now answer a growing share of searches before a person ever sees your listing, and ChatGPT has more than 900 million weekly users who never touch Google for that question at all. A page can rank on page one and still be invisible to every one of them.
This is the audit I run before I recommend a single tactic to a client, and it is the same one I am running across my own site's content as I build out this pillar. Here is exactly how to do it, with no tools required for the first pass and real tool options for when you want it automated.
Quick answer
An AI search visibility audit checks whether ChatGPT, Google AI Overviews, and Perplexity actually mention, recommend, or cite your business when someone asks a question your business should answer. Start manually: ask each AI tool 5 real customer questions and record whether you appear. Then use a free checker (Semrush or Ahrefs both offer one) for a wider scan, and a paid tracker like Otterly AI or Airefs if you want this monitored on a schedule instead of checked by hand. A ranking position in Google does not predict any of this. They are measuring two different things.
By the end of this guide you will have run the manual audit yourself, know which tools actually automate it, and know what a genuinely bad result looks like versus one that just needs more time.
Why This Audit Matters More Than Your Google Ranking Right Now
Google searches in the US ended without a click 68.01% of the time between January and April 2026, up from 60.45% in 2024, according to SparkToro's Similarweb-based study. When an AI Overview appears on the results page, which now happens on more than 20% of Google searches, click-through rates on the organic results below it drop by nearly 60%. Add ChatGPT's 900 million weekly users, most of whom never touch Google for that question at all, and a real share of your potential customers are getting their answer without ever seeing a website, yours or a competitor's.
The platforms do not even source their answers the same way, which is why an audit needs to check all three separately instead of assuming one predicts the others. Research from Profound on citation patterns found ChatGPT leans heavily on encyclopedic sources, Wikipedia alone accounts for 47.9% of its top 10 sources, while Google AI Overviews pull more evenly from Reddit (21.0%) and YouTube (18.8%), and Perplexity favors community platforms even more heavily, with Reddit making up 46.7% of its top citations. Across all three, commercial .com domains still account for 80.41% of ChatGPT's citations, so a real business website is not automatically at a disadvantage against forums and review sites. It still has to earn the citation on its own terms.
What "Showing Up" in AI Search Actually Means
Showing up in AI search is not the same as being indexed by Google. It means an AI system said your business's name, recommended your product, or pulled a specific claim from your page into its answer, without the user clicking through to check.
There are three separate levels here, and most audits confuse them. The first is being retrieved: the AI system's underlying search layer found your page as a candidate source. The second is being cited: the AI system actually pulled text from your page and attributed it, usually with a visible link. The third, and the one that actually drives business results, is being recommended: the AI system names your business specifically as the answer to "who should I use for X," with or without a link attached.
A page can clear the first level constantly and never reach the third. That gap is exactly what this audit is built to find.
The Manual Audit: 5 Questions, No Tools Required
Do this first, before spending money on anything. It takes about 15 minutes and tells you more than most paid dashboards will in the first month, because you are asking the exact questions your actual customers ask, not a generic tracked keyword list.
Open ChatGPT, Perplexity, and Google (to trigger an AI Overview) in three separate tabs and ask each one the same 5 questions:
- The direct name check. "What does [your business name] do?" This confirms whether the AI system has any accurate information about you at all, or whether it is guessing, confusing you with a similarly named business, or has nothing.
- The category question. "Who are the best [your service] providers in [your city or niche]?" This is the one that actually drives leads. Note whether you appear, where in the list, and whether the description next to your name is accurate.
- The comparison question. "Is [your business] better than [a named competitor]?" or "[Competitor A] vs [you], which should I choose?" This shows whether the AI system has enough information to compare you at all, and which way it leans when it does.
- The specific-claim question. Pick one real fact from your own website, a price, a guarantee, a specific process step, and ask a question that should surface it. "How much does [your specific service] cost from [your business]?" If the AI system gets this wrong or has nothing, that page is not being extracted cleanly, whatever its Google ranking says.
- The zero-context question. Ask the category question again, but this time from a fresh chat with no prior context about your industry. This strips out any advantage from earlier turns in the conversation and shows what a first-time user actually sees.
Write down the answer to each one, word for word, in a simple spreadsheet: question, which AI tool, whether you appeared, and what was said about you if you did. That sheet is your baseline. Everything after this section is about improving what is in it.
What a Bad Result Actually Looks Like
A bad result is not "I didn't rank number one." It is one of these four patterns, in order of how often I see them:
| Pattern | What it means | Most likely cause |
|---|---|---|
| You never appear, at all, in any of the 5 questions | The AI system has no confident source to pull from | Thin content, no direct-answer structure, or genuinely low authority |
| You appear in the name check, not the category question | The AI knows you exist but does not associate you with the buying decision | Your content describes what you do, not the specific question a buyer asks |
| You appear, but the details are wrong or outdated | The AI is citing you with low confidence and filling gaps by guessing | Stale content, missing schema, or no visible "last updated" date |
| A competitor appears in your place with weaker content | Structure is beating substance, which is fixable and worth fixing first | Their page states the answer directly in the first two sentences and yours does not |
That last row is the most common gap I find, and the cheapest to fix. It is the exact problem the ANSWER framework in this pillar's AEO guide is built to solve: state the answer in sentence one, not paragraph four.
Automating the Audit: Free and Paid Tools
The manual method above does not scale past a handful of questions, and it does not track change over time. Once you have your baseline, these are the tools that automate the same check across dozens of prompts on a schedule.
Free first pass: both Semrush and Ahrefs publish a free AI visibility checker (search "Semrush AI search visibility checker" or "Ahrefs AI visibility checker" to find the current versions). Run your business name and your main category question through both. They will not track history or monitor changes, but they give you a second, tool-generated data point to compare against your own manual audit.
Paid, ongoing tracking: if the manual audit shows real gaps worth monitoring, a dedicated tracker is the next step. I already compared the real options with current pricing in ChatGPT SEO Tools: 9 Real Options, so I will not repeat the full breakdown here, but the short version: Airefs starts at $24 a month for the tightest budget, Otterly AI at $29 a month covers 15 tracked prompts across 6 AI engines, and Semrush's AI Toolkit at $99 a month makes sense if you are already inside the Semrush ecosystem. All three do the same core job as your manual spreadsheet, just automatically and at a scale a person cannot keep up with by hand. For a full comparison of 9 of these tracking tools with current vendor pricing, see Best AI Search Visibility Tools Compared.
A Real Example From My Own Search Console Data
I am not going to claim a finished AI-citation audit result here that I have not actually run and verified, that would break the same honesty standard this entire framework depends on. What I can show you is the real data point that convinced me this pillar was worth building in the first place.
My own ChatGPT SEO Tools article gets over 1,000 impressions a month in Google Search Console. It sits at position 37, on page 4. That tells me two things clearly: real people are searching for this exact topic, and Google has indexed the page. It tells me nothing about whether ChatGPT or Perplexity are citing it in an answer, because that is a completely separate system with its own retrieval and trust signals, which is the entire point of this article. Running the manual 5-question audit above on that same page, and on the rest of this pillar, is the next real step, and I will publish the honest result once I have run it properly rather than guessing at a number now.
What to Fix First When the Audit Comes Back Bad
Fix in this order, not all at once. Trying to fix everything in one pass usually means nothing gets fixed well.
- Confirm the AI crawlers can actually reach your site first. Check that your robots.txt file is not blocking OAI-SearchBot or the other major AI crawlers, and that the page is genuinely indexed by Google. None of the content fixes below matter if the page was never reachable in the first place.
- Rewrite the opening of your highest-value page next. State the direct answer to your category question in the first two sentences. This is the single content change with the best return for the least effort.
- Add or fix your FAQPage and Article schema. This is the machine-readable version of the same information, and it is one of the first things an AI system's retrieval layer checks before it has to infer anything from unstructured text.
- Name real numbers and specific entities in place of vague claims. "We do great work" gets skipped. "50 of 60 exhibitor stalls booked from one campaign" gets quoted, because there is something specific to lift.
- Re-run the manual 5-question audit in 6 to 12 weeks. That is the realistic window before AI engines re-crawl and re-index a meaningfully changed page, not before.
If your business is a home service company, an ecommerce brand, or a coaching or course business, the vertical-specific playbooks in the AEO pillar guide walk through exactly which questions matter most for your audience before you run this audit again.
How Often to Re-Run This
Once a quarter is the realistic minimum, once a month if you are actively publishing new content or your competitive landscape moves fast. AI engines retrain and re-crawl on schedules none of us control, so a single audit is a snapshot, not a verdict. Treat it the same way you would treat a Google Search Console check: useful once, essential as a habit.
If you want this exact audit run on your own business, with a plan attached for what to fix first, that is the starting point for the AI-driven marketing and SEO work I do at Afraz Alam Marketing Services. Book a call and I will walk through your results with you directly, not hand you a PDF and disappear.
FAQ: AI Search Visibility Audits
What does an AI search visibility audit actually check?
It checks whether ChatGPT, Google AI Overviews, and Perplexity mention, cite, or recommend your business when someone asks a real question in your category, as opposed to checking your Google ranking position, which measures something different.
How do I check if ChatGPT knows about my business?
Ask it directly: "What does [your business name] do?" and "Who are the best [your service] providers in [your city or niche]?" If the answer is wrong, generic, or you do not appear at all, that is your starting baseline.
Are free AI visibility checker tools accurate?
They are useful as a second data point, not a full picture. Free checkers from Semrush and Ahrefs give a single spot-check with no history or trend tracking. A manual audit across the real questions your customers ask, repeated over time, catches more than a one-time free scan.
How is an AI visibility audit different from an SEO audit?
An SEO audit checks technical health, backlinks, and ranking position. An AI visibility audit checks whether an AI system will actually quote, cite, or recommend your business in a generated answer. A page can pass an SEO audit completely and still fail an AI visibility audit if the answer is not stated directly enough to extract.
Does showing up in Google AI Overviews mean I will also show up in ChatGPT?
No. Each platform has its own retrieval and citation logic, and independent research on citation patterns shows real differences between them, ChatGPT leans heavily on encyclopedic sources like Wikipedia, while Google AI Overviews and Perplexity draw more from community platforms like Reddit. Appearing in one does not guarantee appearing in another.
What's a good AI visibility result on this audit?
Appearing accurately in at least 3 of the 5 manual questions across at least 2 of the 3 AI tools is a solid starting benchmark for a small business. Appearing in all 5 across all 3 consistently is rare even for established brands, and is a multi-month goal, not a first-audit expectation.
How long does it take to see a result after fixing what the audit finds?
Realistically 6 to 12 weeks, and longer for competitive categories. AI engines re-crawl and retrain on their own schedules, and there is no guaranteed timeline anyone can honestly promise you.
Want the real numbers on your own AI search visibility, not a guess?
I send real updates on SEO, automation, AI marketing, and paid media (Meta Ads and Google Ads) as they happen, plus case studies of how other businesses grow revenue through performance marketing.
Sign Up for the Newsletter