> For AI agents: a documentation index is available at [llms.txt](https://1alexander.ai/llms.txt).

# What is an AI visibility audit?

An AI visibility audit is a dated record of how named assistants answer a defined set of buyer questions, plus the sources they cite and a plan for what to fix.

[All guides](https://1alexander.ai/articles)

By [Jake Parks](https://1alexander.ai/about)  Updated 16 Nov 2026  5 min read

An AI visibility audit is a dated record of how named assistants answer a defined set of buyer questions, plus the sources they cite and a plan for what to fix. The record should name ChatGPT, Google AI Overviews, Perplexity and Grok, and explain any assistants left out.

In this guide  

## Check the audit contents

Use this acceptance checklist before agreeing to a scope. Ask to see the evidence behind each finding, not just a summary chart.

Scroll sideways to see the full checklist.

| Contents | What you should receive |
| --- | --- |
| Prompt set | Exact buyer questions, why they matter and whether they name \[company\] or ask about a category. |
| Assistants and runs | Named assistants, dates, settings where available, number of questions and repeat runs; record when no answer appears. |
| Mentions versus citations | Separate records for a business being named and a source being linked. Keep referral visits and sales in separate reporting. |
| Who appears instead | Businesses and alternatives named in the same answers, with context rather than a claimed market ranking. |
| Cited sources and URLs | The actual source links, which page supports which statement and any inaccurate facts. |
| Crawler access and indexation | Access checks for relevant pages and evidence of indexation where available; distinguish an access issue from an absent citation. |
| On-site and structured-data fixes | Specific pages and factual corrections, internal links and markup that describes facts already visible to readers. |
| Ranked next steps | Priorities with reasons, owners, dependencies and the proposed page or site change. |
| Retest plan | Questions, assistants, repeat runs and dates to revisit; explain any changes to the sample. |

For Google’s AI features, pages need to be indexed and eligible for a Search snippet. Access checks should include robots.txt and hosting restrictions. Google says there are no extra AI-only schema or file requirements; structured data should match visible facts. Eligibility does not guarantee indexing or inclusion. [\[1\]](https://1alexander.ai/articles/ai-visibility-audit#source-1) Its optimization guide keeps the focus on useful content and SEO fundamentals. [\[2\]](https://1alexander.ai/articles/ai-visibility-audit#source-2)

## Keep the answers readable

A good audit deliverable is dated and includes raw answers, or enough context to reread each observation. You should be able to trace a finding back to a question, assistant, run date and cited URL. For an illustrative question such as “Does \[product\] fit \[company\]’s approval process?”, retain the wording rather than shortening it to “product visibility.”

State the sample size: how many distinct questions, which assistants and how many runs. Explain what was left out, unavailable or changed. Repeated answers for one question do not represent all buyer journeys. A snapshot is not a ranking or a forecast.

Label mentions, citations, referral visits and business outcomes separately. A citation to your page does not establish that someone visited it. A referral visit does not establish a qualified lead or sale. Follow the [measurement guide](https://1alexander.ai/research/measuring-ai-search) for those distinctions, and the [AEO fundamentals guide](https://1alexander.ai/articles/what-is-answer-engine-optimization) for the broader process.

## Turn findings into priorities

Ask what each proposed fix helps a buyer understand. A missing explanation of \[product\]’s integration requirements may call for a clearer page; an incorrect statement needs a factual correction. An absent mention alone does not prove a technical fault. Prioritize useful, supported changes before chasing a chart.

The retest should preserve the original questions and record any changes to wording, coverage or conditions. Compare dated observations and completed fixes, while keeping traffic and sales in their own records. The [visibility-check workflow](https://1alexander.ai/technology/ai-visibility-tracker) explains how those records can inform a managed service.

## Inside a paid engagement

Within 1Alexander’s paid engagement, agreed buyer questions guide dated checks that record the assistant, date, wording, mention and cited page. Where available, the answer is kept for context. The [methodology](https://1alexander.ai/methodology) explains how findings become proposed improvements; [how we work](https://1alexander.ai/how-we-work) explains the handoffs and client approval before pages go live.

The published service starts with a 90-day foundation, followed by monthly updates. Responsibilities, frequency, exclusions and scope belong in the written proposal; check the [current plans](https://1alexander.ai/pricing) before agreeing. The tracker is a workflow within the done-for-you service, not a live dashboard, a free tool or a promise of continuous monitoring across every model. An “AI visibility score” must not be treated as revenue.

When comparing proposals, use the [agency-choice checklist](https://1alexander.ai/articles/how-to-choose-an-aeo-agency) to ask who will make the fixes and how progress will be reported.

## Short FAQ

### What is an AI visibility audit?

It is a dated record of how named assistants answer defined buyer questions, the sources they cite and a plan for what to fix. Its findings apply to the recorded sample.

### What should an AI visibility audit include?

Exact prompts, assistants and runs, mentions and citations, who appears instead, source URLs, access and indexation checks, proposed page fixes, ranked next steps and a retest plan. Keep referral visits and sales separate.

### How can I check how my brand shows up in AI search?

Choose relevant buyer questions, run them in the named assistants and save the question, assistant, date, answer and sources. Record whether your brand is named and repeat the same questions on agreed dates. State the sample and limits.

### How do I measure AEO performance?

Compare dated mentions and citations for a defined question set, record completed improvements, and measure referral visits, qualified leads and sales separately. Visibility alone does not establish revenue or causation.

### What should an AEO checklist include?

For the audit stage, include the contents table above: questions, runs, answer evidence, sources, access checks, proposed fixes and retesting. This is an audit acceptance checklist, not a complete technical implementation checklist.

### What is the difference between a mention and a citation?

A mention names a business or product in an answer. A citation links to a source page. Either can occur without the other, and neither establishes a referral visit or sale.

## Sources

Google guidance checked on 16 November 2026. The checklist is a practical buying framework, not a Google certification or a claim of proprietary research. Service details come from the linked 1Alexander methodology, how-we-work, tracker and pricing pages.

\[1\] [Google Search Central, AI features and your website](https://developers.google.com/search/docs/appearance/ai-features)

\[2\] [Google Search Central, Optimizing your website for generative AI features on Google Search](https://developers.google.com/search/docs/fundamentals/ai-optimization-guide)

## Related reading

[Choose an AEO agency](https://1alexander.ai/articles/how-to-choose-an-aeo-agency)

## Put the guide into practice

[See the visibility-check workflow](https://1alexander.ai/technology/ai-visibility-tracker)
