Home/AI SEO Services/AI Visibility Audit
Measured on Live AI Answers

AI Visibility Audit

Your brand checked on 12 editable prompts across eight AI surfaces, with citations and mentions scored apart.

Andy Wang · founder, does the work on every accountB.S. Pharmacology, UC Santa Barbara · Pokémon, Disney, Mastercard, Mailchimp · About Andy →

An AI visibility audit measures how often AI search engines cite a brand’s site and name the brand in their answers. SERP Masters is an SEO agency that runs this audit for businesses. The free run sends 12 editable prompts to eight AI surfaces. It scores four separate panels: citation presence, mention presence, retrieval access, and extractability plus entity. The result is a per-engine baseline of where AI answers send buyers, shown on screen before any email is requested.

Request the free AI Visibility Audit
Measured, per enginePortfolio
15,332AI citations
counted once
+80%client AI citations
year over year
5 of 6zero-citation
sites now cited
Google AI Overviews and AI Mode
ChatGPT and Perplexity
Gemini, Copilot and Grok
Google Search Console for clicks

Citation figures are de-duplicated across Ahrefs, DataForSEO and live renders. Ranking and citation outcomes cannot be projected with certainty.

15,332AI citations across the client portfolio
700+Google Business Profiles managed
$1M+Managed link building spend

Buyers ask AI engines for shortlists. ChatGPT, Perplexity, Gemini, Claude, Copilot and Google AI Overviews answer with a few names and links. A rank tracker cannot tell you whether those answers cite your site and name your brand.

The SERP Masters audit shows you the results engine by engine, and its prompt-by-prompt table traces every number to an answer you can read.

The measurement is a citation-presence rate on a fixed prompt panel, scored per engine and never blended. SERP Masters uses the same measurement on client work, where AI citations grew 80% year over year.

The audit is the Diagnose step of the AI SEO services engagement at SERP Masters, so nothing gets fixed before it is measured. You need a domain, a brand name and a location to start.

What does the SERP Masters AI visibility audit measure?

The SERP Masters AI visibility audit measures four things, each reported as its own panel.

PanelQuestion it answersHow it is counted
Citation presenceDoes the AI answer link to your site as a source?Share of readable prompt and surface pairs with your domain as a linked source
Mention presenceDoes the AI answer name your brand?Share of prompt and surface pairs with your brand named in the answer text
Retrieval accessCan the bots that fetch pages for AI answers reach you?Of 16 AI user agents tested, the retrieval and grounding bots that receive a normal page
Extractability and entityCan an engine lift a clean answer and know who you are?9 page checks (rendering, answer capsule, schema / structured data) and 6 entity checks

Citation presence

Citation presence counts only answers with readable sources. An AI Overview with no readable source list is recorded and left out of the denominator, so a hidden list never counts against you.

Mention presence

Mention presence matters because a named brand makes the buyer’s shortlist even without a link. A brand named more often than it is linked is known to the engine, which has no page of yours it wants to cite.

Retrieval access

Retrieval access is tested by requesting your site as 16 AI user agents, among them OAI-SearchBot, ChatGPT-User, PerplexityBot, ClaudeBot, Googlebot and Bingbot. Only the 11 that fetch pages to answer a live question count. Training-only crawlers such as GPTBot and CCBot appear in the access matrix and stay out of the score, since blocking them does not stop a live answer from citing you.

Extractability and entity

Extractability and entity checks test whether an engine can quote a usable passage and tie it to the right organization. Page checks compare raw HTML with the rendered page, confirm structured data survives rendering, and look for an answer capsule near the top. Entity checks look for an Organization node with a stable @id, sameAs links and consistent contact details.

Why citations and mentions are never blended

Citations and mentions call for different fixes. A brand cited on 40% of prompts and named on 10% has a different problem from one named on 40% and cited on 10%; an average makes them identical. Engines are kept apart for the same reason: a brand can hold steady in Perplexity and sit at zero in Gemini.

Which eight AI surfaces does the audit check?

The audit checks eight AI surfaces: five answered live during your run, and three read from the Ahrefs AI citation index, kept in their own column.

SurfaceSource in the auditHow the answer is collected
ChatGPTLiveWeb search on; answer and source links captured
PerplexityLiveAnswer and cited sources captured
GeminiLiveGoogle Search grounding; answer and sources captured
ClaudeLiveWeb search on; answer and sources captured
Google AI OverviewsLiveOne results page per prompt at your location, marked present, absent, or present without readable sources
Google AI ModeAhrefs indexStored citation counts for your domain
Microsoft CopilotAhrefs indexStored citation counts for your domain
GrokAhrefs indexStored citation counts for your domain

The guide to earning AI Overview citations covers what AI Overviews draw from.

The 12 prompts: shown and editable before the run

All 12 prompts are on screen before the run, and you can rewrite any of them to match what your buyers ask.

01Four commercial prompts, such as “best [service] in [city]” or “best [category] for [use case]”.
02Two brand promptsthat check accuracy.
03Four informational promptsa buyer asks while learning about the problem.
04Two comparison promptsagainst an alternative buyers already weigh.

Brand prompts alone flatter a brand: a business named on every brand prompt can be missing from every commercial prompt, where the shortlist forms. Each prompt is also classed local or national.

What does the AI visibility report show?

The AI visibility report shows the four panels, a Judged grade, and the answer behind every number.

Grade and panels:the A to E grade, labeled Judged, beside the four measured panels it reads
Surface table:one row per engine
Cited-domain league:the domains AI answers cite most, with your position
Recommended-instead league:brands named where yours was absent, with each competitor’s share of voice
Access matrix and checklist:16 agents, 15 checks
Ordered fix list:what to change first
Prompt-by-prompt table:every prompt, surface and answer
Method note:what was counted or excluded, and why

The report shares by link. Email is requested only for a PDF copy or a full-audit request, pre-filled with your run ID.

How does the free audit work?

The free run takes six steps and about a minute.

01Enter details:domain, brand, location
02Review the prompts:edit any of the 12, add competitors for the leagues, then confirm
03Access test:16 AI user agents request your site
04Page and entity checks:your key page is fetched raw and rendered
05Answer collection:five live surfaces answer, three index surfaces are read
06Report:panels, grade, leagues and fix list appear here

One limit: the free run sends each prompt once, and AI answers vary between runs, so the result is a snapshot.

In what order should you fix AI visibility problems?

Fix AI visibility problems in dependency order: access, then the page, then the gaps in the answers. A lower fix does little while one above it is open.

01Access blocks.A bot that gets a 403, a challenge page or an empty shell cannot cite the site. Firewall and CDN rules are the usual cause, and technical SEO work on crawler access clears them.
02Page and identity gaps.A missing answer capsule, JavaScript-only content or absent Organization schema leaves the engine less to lift.
03Named more than linked.The fix is a page worth citing, built for those prompts.
04Competitor set.The two leagues show who takes your place and where corroborating mentions would count.
05Partial answers.Incomplete or wrong descriptions go last, with the correct facts beside each.

How does a measured audit differ from a modeled AI visibility score?

A measured audit reads real AI answers. A modeled AI visibility score estimates visibility from homepage signals, as many of the tools SERP Masters reviewed do.

QuestionModeled AI visibility scoreSERP Masters AI visibility audit
Where the number comes fromHomepage signals scored against a checklistLive answers from five surfaces plus three index surfaces
PromptsHidden or generated without review12 prompts shown and editable
Citations and mentionsOften one combined scoreTwo separate panels
Bot accessrobots.txt / AI crawlers read, if checked at all16 user agents requested, retrieval and training bots separated
GradePresented as a measurementLabeled Judged, shown next to the panels
Answer variationNot mentionedStated in the method note; three runs per prompt in the full audit

This audit is a point-in-time measurement. Ongoing monitoring is a separate job, covered in the comparison of AI visibility tools and trackers.

Free AI visibility checker or full audit?

The free AI visibility checker gives a 12-prompt baseline. The full audit covers a whole topic cluster, runs every prompt three times, and is quoted from your free run.

ItemFree AI visibility checkerFull AI visibility audit
Prompts12, editable40 to 60 per topic cluster
Runs per promptOneThree, reported with a band
Bot access16 agents from one locationPer-bot fetch matrix through your CDN
Cited pages / sourcesCited-domain leagueSource map with exact cited URLs
CompetitorsNamed in the leaguesBenchmarked, with gap lists per competitor
DeliverableOn-page report, share link, PDF on requestWritten report and workbook
TurnaroundAbout a minuteFive business days
PriceFree\\quoted after the free audit

The full audit is priced by scope: the number of topic clusters, competitors benchmarked and locations the prompts run in. It also depends on you, since the per-bot matrix needs access to your CDN settings.

Who is the AI visibility audit for?

The audit is for businesses whose buyers ask AI engines for recommendations before they call or book.

B2B and SaaS teamsthat see competitors in ChatGPT shortlists.
Law firms and healthcare providers, where an inaccurate AI description carries more risk than an absent one.
Ecommerce brandswhose comparison prompts return marketplaces and review sites.
Local service businesseschecking whether AI answers name them in their city.
Marketing and SEO managerswho need per-surface numbers for leadership.

The audit also works as an LLM visibility checker. Its live surfaces run with web search on, so it measures what an engine retrieves today; the guide to how LLMs represent brands covers what a model repeats from training data.

What results has this method produced?

The method this audit starts has produced measured AI citation growth across the SERP Masters client portfolio.

More client results are on the SERP Masters case studies page.

From a SERP Masters client
“After a year of paying an agency that quietly did nothing, working with SERP Masters was like someone flipping the lights on. Andy audited the damage, gave me the honest version of how bad it was, and got us moving again.”
A SERP Masters client, quoted with consent

What happens after the audit?

The audit is the Diagnose step of the SERP Masters method: Diagnose, Architect, Execute, Corroborate, Measure. The fix list feeds Architect and Execute, where each prompt class gets an owning page. Direct-answer pages follow direct-answer AEO work; pages in multi-source answers follow GEO work for synthesized answers. The cited-domain league sets Corroborate targets, and Measure re-runs the same prompts on the same surfaces.

Report terms such as grounding and query fan-out are defined in the glossary of AI search terms.

Who runs the SERP Masters AI visibility audit?

Andy Wang, founder of SERP Masters, designed the audit and reviews every full audit. He holds a B.S. in Pharmacology from UC Santa Barbara and has run SEO work for Pokémon, Disney, Mastercard, Mailchimp, US Soccer, Clorox, Wynn Las Vegas and BigCommerce.

The person who reads your full audit is the one who would act on it. More on Andy Wang is on the About SERP Masters page.

From a SERP Masters client
“What I value most is that Andy tells me the truth even when it’s not what I want to hear. Across the entire engagement I never once felt managed or spun. That’s rare in this business, rarer than it should be.”
A SERP Masters client, quoted with consent
AI Visibility Audit FAQ

Answered the way an engine would extract it.

Each answer opens with the answer. The FAQPage schema is generated from these fields and nothing else.

01What is an AI visibility audit?

An AI visibility audit is a measured check of whether AI search engines cite a brand’s site, name the brand, can fetch its pages, and can quote a usable passage. The SERP Masters version runs 12 editable prompts across eight AI surfaces and reports those four results as separate panels, each traced to the answers behind it.

02How do you measure AI visibility?

You measure AI visibility by sending a fixed set of buyer prompts to each AI surface and recording whether each answer cites the brand and whether it names it. SERP Masters reports both as per-surface rates, adds a bot access test and page checks, and keeps the prompt set fixed so later runs compare like for like.

03What does AI visibility mean?

AI visibility means how often a brand appears in answers from AI engines such as ChatGPT, Perplexity, Gemini, Claude and Google AI Overviews. It covers two outcomes: being cited as a linked source and being named in the answer text. SERP Masters measures them separately because a brand that is named but never cited needs a different fix.

04Can ChatGPT do an SEO audit?

ChatGPT can review a pasted page and suggest changes, and it can answer the prompts an audit tests. It cannot request your site as 16 AI user agents, compare raw and rendered HTML, or score its answers against other engines. The SERP Masters audit treats ChatGPT as one of eight surfaces it measures.

05Is the AI visibility checker free?

Yes, the SERP Masters AI visibility checker is free for one run of 12 prompts across eight surfaces, and results appear on the page without an email address. One-time audits and project work are quoted after the free audit.

06How is the AI visibility score calculated?

The SERP Masters AI visibility score is an A to E grade labeled Judged, set from four measured panels: citation presence, mention presence, retrieval access, and extractability plus entity. The weights stay internal. The four panels always sit beside the grade, so you never have to take the grade on trust.

07Why do AI visibility results change between runs?

AI engines can return different answers to the same prompt minutes apart, and Google AI Overviews also change by location. The free run is a single snapshot and says so in its method note. The full audit sends each prompt three times and reports a band, which separates a real gap from ordinary answer variation.

Andy Wang, SEO consultant and founder of SERP Masters
Who runs the audit

Andy Wang, founder of SERP Masters.

Andy Wang holds a B.S. in Pharmacology from UC Santa Barbara and has run SEO work for Pokémon, Disney, Mastercard, Mailchimp, US Soccer, Clorox, Wynn Las Vegas and BigCommerce. He reads every audit result himself and replies with the citation gaps ranked by commercial value.

Related case studies AI citations, portfolio15,332 Telehealth AEO and SEO1,100+ Ecommerce AI searchAIO cited Every count on this page links to the page that proves it.
AI VISIBILITY AUDIT ·AI VISIBILITY AUDIT ·AI VISIBILITY AUDIT ·AI VISIBILITY AUDIT ·
Free AI Visibility Audit

Run the free AI visibility audit

The free audit shows which AI engines cite your site and name your brand, which bots can reach you, and what to fix first. Enter a domain, edit the 12 prompts, and read the per-engine report on this page. If the report shows a gap worth closing, SERP Masters quotes the full audit from your run, Andy Wang reviews it himself, and any follow-on work runs month to month.

Request the free audit