Search Standing/AI visibility audit
AI visibility audit

An AI visibility audit counts ChatGPT, Gemini, AI Mode and AI Overviews separately.

An AI visibility audit, also called a GEO audit, measures whether ChatGPT, Gemini, Google AI Mode and Google AI Overviews name your company for your category, each counted separately, with the sources each one drew on and the fault behind every gap. One company we measured in August 2026 had 44 mentions in ChatGPT and 2 in AI Overviews in the same week. A blended score would have hidden it.

Five steps, the findings they produce, and what you receive. The first analysis is free.

4Engines, counted separately
10Crawlers checked for access
44 vs 2ChatGPT vs AI Overview, one company, Aug 2026
What it is[1]

What is an AI visibility audit?

An AI visibility audit measures whether and how often ChatGPT, Gemini, Google AI Mode and Google AI Overviews name or cite a company when asked a question in its category and market. It counts mentions per engine, maps which sources each engine draws on, checks that the company's site is retrievable by each engine's crawler, confirms the business resolves to one entity, and explains the gaps between engines. It ends with a thesis, not a score.

The five steps of an AI visibility auditCrawler access, entity resolution, four-engine mention count, cited-source map, and the gap between engines, in that order. 01Crawler accessROBOTS.TXT AND SERVER RESPONSE PER BOT02Entity resolutionONE NAME, ONE PLACE, EVERYWHERE?03Four-engine countCHATGPT, GEMINI, AI MODE, AI OVERVIEW, SEPARATELY04Cited-source mapWHICH PAGES EACH ENGINE DRAWS ON05The gapWHERE THE ENGINES DISAGREE, AND WHY
Five steps. The fifth is the finding: the engines disagree, and the disagreement names the fault.

The order matters. Crawler access is checked first because a site that refuses a bot cannot be cited by it, and the finding reframes everything after. Entity resolution is second because an engine that cannot place a business does not name it. Only then are the four engines queried, and the cited sources mapped, so that the gaps in step five have an explanation rather than a shrug.

What it finds[2]

What does an AI visibility audit usually find?

That the engines disagree, and that the disagreement is diagnostic. One recurring pattern in our 2026 analyses is strong ChatGPT and AI Mode counts beside a near-zero AI Overview count, caused by a server refusing Googlebot. Another is a company absent from all four because its name, address and category differ across the web. A third is a company cited for a topic it no longer wants, such as a neighborhood it moved away from.

One company across four engines, August 2026Mentions, cited pages and the likely cause of the gap for one commercial printer measured in one week. CHATGPTGEMINIAI MODEAI OVERVIEWMentions4416432Own pages cited93110Top cited sourceTheir homepageYelpTheir homepageA competitorReadsCache + searchLive indexFan-out searchLive pageWhy it failedHTTP 406 to crawler
A blended score would have averaged this to a healthy number. The fourth column is a developer fix.
Crawler refusedHTTP 406, 429 or a bot challenge to Googlebot or OAI-SearchBot. One engine goes dark while the others are fine.
Unresolved entityThree address formats, two names, a mailing address in another town. No engine can place the company, so none names it.
Named for the wrong thingTop AI topic for one New York printer, measured August 2026, was the neighborhood it left. The web still said it was there.
Competitors as your sourcesTop cited sources for one industrial company, measured October 2026, were three competitors. The engine learned the category from them.
Absent from the roundupsThe directories and listicles ChatGPT cites for the category do not include the company. Nothing on its own site can fix that.
Blocking the wrong botGPTBot blocked to stop training, OAI-SearchBot blocked by the same rule, ChatGPT search citations gone.
Tool versus audit[3]

How is this different from an AI visibility tracking tool?

A tracking tool produces a blended visibility score and a trend line. An audit produces four counts, the sources behind each, the fault behind each gap, and the fix. Tools are useful once the faults are known; they are not how the faults are found. A server refusing one crawler, one of the faults we find most, is invisible to a dashboard because the dashboard averages the engine that failed with the three that did not.

A tool dashboard compared with an auditA dashboard gives a blended score and a trend line. An audit names the engine, the fault, the source and the fix. VISIBILITY DASHBOARDTHIS AUDITOne blended scoreFour counts, one per engineA trend lineThe cited sources, namedMentions, undifferentiatedMentions, by engine, with the queryNo causeThe fault behind the gapSold monthlyDelivered once, repeated quarterly if wantedReads your siteReads your site, your robots.txt, your server, and the web
A dashboard tells you a number moved. An audit tells you why, and what to change.

The companies selling visibility scores are selling a reasonable product to the wrong stage. Monitoring belongs after the diagnosis, the specification and the pages. Bought first, a score tells a company it has a problem it already suspected, in a form that cannot say where.

The deliverable[4]

What do you receive from an AI visibility audit?

A written document: crawler-access results for every major bot, the entity check with every variant found, four mention counts with the queries used, the cited-source map per engine, the gap analysis explaining each disagreement, and a thesis stating what is wrong and what fixes it. The free first analysis covers a four-engine mention count, a crawler-access check, an entity check, and the thesis, in summary; the paid audit carries the full measured data, the cited-source map and the gap analysis.

  1. Crawler access tableGooglebot, Google-Extended, GPTBot, OAI-SearchBot, ChatGPT-User, ClaudeBot, Claude-SearchBot, PerplexityBot, Bytespider, CCBot: robots.txt rule and live server response for each.
  2. Entity checkEvery name, address and phone variant found on the site, Google Business Profile, and the directories the engines cite. Mismatches listed.
  3. Four mention countsChatGPT, Gemini, AI Mode, AI Overview, each queried the same ten ways for the category and market (ten phrasings of the buyer's question). Counts and the queries.
  4. Cited-source mapWhich pages each engine drew on, classified: own site, competitor, directory, roundup, review, press.
  5. Gap analysisEach disagreement between engines, with the measured cause.
  6. The thesisOne paragraph: what is wrong, what it costs, what fixes it. Verifiable in ten minutes.

The AI visibility audit is run inside every B2B SEO audit as steps four and five, and is the part most SEO audits leave out. It can also be run alone for a company that already ranks well and wants to know why AI does not name it.

Cost[5]

How much does an AI visibility audit cost?

The first analysis is free and contains a four-engine mention count, a crawler-access check, an entity check, and the thesis, in summary. The full audit with measured data is included in every engagement, from $20,000 and is not sold separately.

WhatCostContains
Free analysisNothingFour-engine mention count, crawler-access check, entity check, thesis, in summary. Two to three pages.
Full engagementFrom $20,000The paid audit with measured data, a specification for four pages, the pages written.

Full detail on the pricing page. The comparison of AI visibility audits sets this beside what other providers include, and the version for manufacturers covers how manufacturing buyers ask.

Questions[6]

Questions about the AI visibility audit

What is an AI visibility audit?
An AI visibility audit measures whether ChatGPT, Gemini, Google AI Mode and Google AI Overviews name or cite a company for its category and market. It checks crawler access for each engine, confirms the business resolves to one entity, counts mentions per engine with the queries used, maps the sources each engine cites, and explains the gaps between engines. It ends with a thesis rather than a score.
What is a GEO audit?
A GEO audit, or generative engine optimization audit, is another name for an AI visibility audit: a measurement of whether AI engines such as ChatGPT, Gemini, Google AI Mode and Google AI Overviews name a company for its category and market, with crawler access, entity consistency, cited sources and the gaps between engines checked. The two terms describe the same work.
How is AI visibility measured?
By asking each engine the question a buyer would ask, with the category and market, ten ways, and recording which companies it names and which sources it cites. Each engine is counted separately. Blended scores are avoided because the engines disagree and the disagreement is where the faults are.
Why do AI engines give different results for the same company?
Because they read different things. Google AI Overview needs the live page and fails if the server refuses Googlebot. ChatGPT draws on training data and third-party sources and can name a company whose site is unreachable. AI Mode runs several searches per question and can also draw on cached sources. Gemini reads Google's index but answers in a chat, so its results often fall between the Overview and ChatGPT. One company measured August 2026 had 44 ChatGPT mentions and 2 AI Overview appearances in one week; the cause was an HTTP 406 returned to Google's crawler.
What is the difference between an AI visibility audit and an AI visibility tracking tool?
A tracking tool produces a blended score and a trend line. An audit produces four counts, the sources behind each, the fault behind each gap, and the fix. A server refusing one crawler, one of the faults we find most, is invisible to a dashboard that averages the failed engine with the three that worked.
How much does an AI visibility audit cost?
The first analysis is free and contains a four-engine mention count, a crawler-access check, an entity check, and the thesis, in summary. The full audit with measured data, the cited-source map and the gap analysis is included in every engagement, from $20,000.
Which AI crawlers should a B2B website allow?
All of the major ones, if the company sells to buyers who use AI: Googlebot and Google-Extended, GPTBot, OAI-SearchBot and ChatGPT-User, ClaudeBot and Claude-SearchBot, PerplexityBot, and the rest. Blocking a search crawler removes the company from that engine's citations. The audit checks robots.txt and the live server response for each.
Sources[S]

Sources and measurement notes

  1. Google Search Central: Google crawlers and user agents https://developers.google.com/search/docs/crawling-indexing/overview-google-crawlers
  2. OpenAI: overview of OpenAI crawlers https://platform.openai.com/docs/bots
  3. Anthropic: web crawling and site owner controls https://support.anthropic.com/en/articles/8896518-does-anthropic-crawl-data-from-the-web-and-how-can-site-owners-block-the-crawler
  4. Search Standing four-engine measurement, August 2026 https://searchstanding.com/ai-search-statistics/

Figures described as measured are from Search Standing's own analyses using Semrush, Google Search Console and direct queries to each AI engine, on the dates stated. Published ranges move quarter to quarter; the patterns are what we rely on.

Start[8]

Start with the diagnosis.

Send your company and your market. We will run the analysis and send back a summary naming your competitors, the terms they are winning, and where you sit against them.

No call required and no obligation. Read it, check it yourself, and decide. If the findings are worth discussing after that, we can talk then.

Get your 2–3 page analysis

Prepared for your company against the competitors you name. Written by the person who would run the paid work.

•Back within three working days  ·  No charge
Send it. No charge

Or email contact@searchstanding.com directly.

What arrivesTwo to three pages, prepared for your company.
WhenWithin three working days.
What it costsNothing. No call, no obligation.
What it is notA template, a tool report, or a sales deck.