How to Run an AI Visibility Audit (Free Checklist Included)

An AI visibility audit tells you exactly where your brand appears across generative engines like ChatGPT, Perplexity, and Google AI Overviews, where it is invisible, and where it is being misrepresented. Without this baseline, any effort to improve your AI search presence is guesswork. This guide walks you through the complete process, from gathering your tools to tracking measurable improvements over time, with a free checklist at the end to keep every step organized.

Most brands running their first GEO audit discover they are invisible in the majority of category-relevant prompts. That gap represents real buying decisions happening without your brand in the room. The steps below give you a structured way to close it.

What you need before starting the audit

Before you run a single prompt, define the scope of your audit. A GEO audit measures brand presence across AI engines by answering three questions: where are you visible, where are you invisible, and where is the information about you inaccurate or outdated? Knowing which platforms you are auditing and what success looks like keeps the process focused and repeatable.

Gather the following before you begin:

• AI platforms to audit: ChatGPT, Perplexity, Google AI Overviews, Claude, and Gemini. These five cover the majority of generative AI search activity in 2026.
• A list of 10 to 30 “money prompts”: Conversational questions a buyer in your category would type into ChatGPT or Perplexity. These replace keywords as the core measurement unit in a GEO audit.
• A competitor shortlist: Identify three to five direct competitors you will track alongside your own brand.
• Tracking tools: At minimum, a spreadsheet. Optionally, tools like Ahrefs Brand Radar or Semrush AI Toolkit to automate prompt monitoring at scale.
• Technical access: Your robots.txt file, your CMS (WordPress or equivalent), and your schema markup setup.

A first-pass audit for a standard business website typically takes two to three weeks of focused work. Plan for it: rushing the process produces a snapshot with gaps rather than an actionable framework. Once your tools and prompt list are ready, move to the first active step.

Check how generative engines currently cite your brand

The first active step is establishing a citation baseline: how often do AI engines mention your brand, in what context, and alongside which competitors? This is the core of your AI search audit, and it requires manual testing to start.

1. Open ChatGPT in a clean incognito session (logged out, no prior conversation history). This prevents prior context from shaping the responses and gives you a true baseline.
2. Run each of your money prompts one by one. Record whether your brand appears, what position it occupies in the response, what competitors are mentioned, and whether your domain is cited as a source.
3. Repeat each prompt three to five times across different sessions. AI responses vary, so a single run is not representative.
4. Repeat the full prompt set in Perplexity. Perplexity shows numbered inline citations you can scan directly, making it easier to identify which specific URLs are being pulled.
5. Run the same prompts in Google AI Overviews and note citation sources there separately, as the source pool differs from conversational AI platforms.
6. For each competitor on your shortlist, run the same prompts and record their citation rates alongside yours.

After this step, you should have a populated spreadsheet showing your brand’s AI visibility rate across each platform and a clear picture of which competitors are consistently cited where you are not. Ahrefs Brand Radar can surface queries where competitors appear but your brand does not, which is particularly useful for identifying the highest-priority gaps. Approximately 80% of URLs cited by AI platforms do not rank in Google’s top 100 results for the same query, so do not assume your Google rankings predict your AI citation rate. They often do not.

Audit your content for GEO readiness

GEO-ready content is structured so that AI retrieval systems can extract and reproduce it accurately. The structural signals that matter here are different from traditional SEO. AI engines parse content like a data feed: clear hierarchy, self-contained answers, and named entities all increase the probability of citation.

Check your page structure

Review your highest-traffic pages first. These already carry search engine trust, making them the strongest candidates for GEO optimization. For each page, verify the following:

• A unique H1 that names the primary topic or entity clearly
• H2 and H3 headings framed as questions a user would ask, with a direct answer in the first sentence below each heading
• Short, self-contained paragraphs of roughly 50 to 150 words each
• Lists used to present options, steps, or comparisons (the majority of pages cited by AI engines use them)
• A structured FAQ section where applicable

Check your entity signals and content freshness

AI engines use entity recognition to understand what a page is about and whether it is authoritative. Establish one primary entity per page and support it with three to six related entities named explicitly in the text. Generic references weaken this signal: name tools, people, standards, and platforms by their actual names.

Content freshness also affects citation probability. AI retrieval systems prefer recent content, so pages that have not been updated in more than six months are at a disadvantage. Add a quarterly review to your workflow: update statistics, verify facts, and refresh examples on all strategic pages at least every three months.

One structural finding worth acting on immediately: content placed in the first 30% of a page receives the majority of AI citations. Lead every page with its most important claim, not with background context or a narrative build-up.

Evaluate your technical and authority signals

Technical issues can silently block all your content work. Even a perfectly structured, entity-rich page will not be cited if AI crawlers cannot access it. This step checks the infrastructure that determines whether generative engines can read your site at all.

Verify AI crawler access

Check your robots.txt file for any rules that block GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, or Google-Extended. These are the primary crawlers used by ChatGPT, Claude, Perplexity, and Google AI Overviews, respectively. If any are blocked, AI engines stop citing your pages regardless of content quality.

If your site runs behind Cloudflare, check your Cloudflare security settings specifically. Since mid-2025, Cloudflare has blocked all known AI crawlers by default on new domains. The “Block AI bots” rule does not distinguish between training crawlers and retrieval crawlers, so enabling it prevents citation entirely. In your Cloudflare dashboard, navigate to the AI Crawl Control panel and set GPTBot, PerplexityBot, and Google-Extended to Allow.

Audit your schema markup

Schema markup is one of the highest-impact technical signals for AI citation. Prioritize three schema types:

• FAQPage schema: AI engines extract Q&A pairs directly from FAQPage markup. Any page with a FAQ section should have this implemented.
• Article or BlogPosting schema: Include headline, description, author, datePublished, and dateModified as a minimum. These properties help AI engines assess authorship and recency.
• Speakable schema: Microsoft Copilot actively uses Speakable markup to identify content for voice-based AI responses.

Assess your authority signals

AI engines use E-E-A-T signals (Experience, Expertise, Authoritativeness, Trustworthiness) as a gating filter, not just a ranking nudge. Pages that fail this filter are excluded from citations entirely, not ranked lower. Check that author bios are present and link to verifiable external profiles, that your content cites named sources, and that your brand is discussed in third-party editorial coverage, industry forums, and review platforms. GEO audit research consistently shows that earned media and external mentions teach AI models to recognize a brand as credible, independent of the content published on the brand’s own site.

Score and prioritize your audit findings

Raw audit findings are only useful if they are organized into a prioritized action list. A structured scoring framework turns your observations into something you can track over time and share with your leadership team.

Score your audit across five dimensions, each rated from 0 to 100:

1. AI citation rate: The percentage of your money prompts where your brand appears across the platforms you audited.
2. Content structure quality: How well your pages follow GEO-ready structure (H1-H2-H3 hierarchy, self-contained paragraphs, FAQ sections, entity density).
3. Technical accessibility: Whether AI crawlers can access your pages and whether schema markup is correctly implemented.
4. Authority and E-E-A-T signals: Author credibility, external mentions, and earned media presence.
5. Competitive position: Your citation rate relative to the three to five competitors you tracked in Step 2.

A score of 70 or above across all five dimensions indicates effective AI visibility. Most brands running their first audit land between 25 and 45. Scores below 60 signal high AI invisibility risk and should be addressed before anything else. Use an Impact vs. Effort matrix to sequence your fixes: high-impact, low-effort actions go first. With your scorecard complete, you have a gap map that directly informs what to fix, what content to create, and what earned media to pursue.

Fix the highest-impact gaps first

With your scored gap map in hand, work through fixes in order of impact and effort. The sequence below reflects what moves the needle fastest for most brands in the first four weeks.

Week 1 to 2: Technical and structural fixes

1. Update robots.txt to explicitly allow GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, and Google-Extended.
2. If using Cloudflare, adjust the AI Crawl Control settings as described in the previous step.
3. Implement FAQPage schema on any page that contains a FAQ section. Use Google’s Rich Results Test to validate the markup.
4. Add Article or BlogPosting schema to your key content pages, including author, datePublished, and dateModified fields.
5. Restructure your highest-traffic pages to use H1-H2-H3 hierarchy with question-based headings and direct answers in the opening sentence of each section.
6. Add structured FAQ sections to product and service pages where they are missing.

Week 3 to 4: Content and authority fixes

Begin with your top five to ten pages by organic traffic. These pages already have search engine trust, so GEO optimization here compounds an existing advantage. Add named entities, update outdated statistics, and ensure the first 30% of each page contains the core claim or answer.

Earned media is the fix most brands skip, and it is among the most effective. Digital Bloom’s citation analysis found that distributing content across third-party news sites produced a significant median lift in AI citations. Invest in presence on Reddit threads, LinkedIn, and niche industry forums relevant to your category. Wire-distributed press releases also produce measurable citation growth over a three- to six-month window. Only a small fraction of GEO practitioners currently prioritize digital PR, which means this is one of the widest gaps between evidence and adoption in the discipline.

If you want to accelerate this process without building a new internal workflow, AI visibility optimization through a hybrid model like WP SEO AI handles the technical restructuring and content optimization inside WordPress while specialists manage the earned media and authority-building strategy in parallel.

Track AI visibility improvements over time

A GEO audit is not a one-time task. AI retrieval systems evolve continuously, and your competitive position changes as other brands optimize. Tracking the right metrics on a consistent cadence is what turns a single audit into an ongoing growth program.

The metrics that matter

Track six core metrics across your fixed prompt set:

• Visibility score: The percentage of tracked prompts where your brand appears.
• AI Share of Voice: The percentage of AI-generated responses in your category that mention your brand. This is the primary KPI for GEO programs in 2026. If 100 relevant AI answers are generated and your brand appears in 28, your AI Share of Voice is 28%.
• Citation share: How often your domain URLs are cited as sources, not just your brand name is mentioned.
• Competitor citation rate: How your visibility score compares to the competitors you tracked in Step 2.
• Sentiment: Whether AI engines describe your brand accurately and positively.
• Positioning accuracy: Whether the AI’s description of your products or services matches reality.

Tracking cadence

Run your full six-area audit quarterly. AI retrieval systems evolve fast enough that quarterly is the minimum useful cadence for a comprehensive review. Monthly, run your 10 to 15 priority money prompts manually across ChatGPT, Perplexity, and Google AI Overviews to catch significant changes between full audits. Trigger an immediate re-audit after any major CMS migration, URL structure change, or template revision, as these can silently break schema and entity signals without triggering any obvious errors.

Most brands begin seeing measurable improvements in AI citation rates within 60 to 90 days of implementing the fixes from this audit. Significant Share of Voice growth typically compounds over four to six months as AI systems build topical authority associations with your brand. Use the scorecard from Step 5 as your baseline and reassess each dimension at every quarterly audit to track progress accurately.

Every metric in your tracking set is prompt-dependent. Citation rate, Share of Voice, and sentiment scores are only comparable when measured against the same fixed prompt panel. Do not benchmark your figures against numbers from a different company using a different prompt set. Benchmark against your own trend line and your named competitor set.

Free AI visibility audit checklist

Use this checklist to run your audit systematically. Check each item off as you complete it.

Before you start

• Define the five AI platforms you will audit (ChatGPT, Perplexity, Google AI Overviews, Claude, Gemini)
• Build a list of 10 to 30 money prompts a buyer in your category would use
• Identify three to five competitors to track alongside your brand
• Set up a tracking spreadsheet with columns for platform, prompt, brand mentioned (yes/no), position, competitors mentioned, and sources cited

Citation baseline

• Run all money prompts in ChatGPT (incognito, logged out), three to five times each
• Run all money prompts in Perplexity and record cited URLs
• Run all money prompts in Google AI Overviews
• Record competitor citation rates for the same prompt set

Content structure

• Unique H1 on every key page
• H2/H3 headings framed as questions with direct answers in the opening sentence
• Self-contained paragraphs of 50 to 150 words
• Lists used on product and service pages
• FAQ section present on key pages
• Core claim placed in the first 30% of each page
• Named entities (tools, people, platforms, standards) used throughout
• All strategic pages updated within the last six months

Technical access

• robots.txt allows GPTBot, ClaudeBot, PerplexityBot, OAI-SearchBot, and Google-Extended
• Cloudflare AI Crawl Control set to Allow for key crawlers (if applicable)
• FAQPage schema implemented and validated on FAQ pages
• Article or BlogPosting schema on key content pages (with author, datePublished, dateModified)
• Schema validated using Google’s Rich Results Test

Authority signals

• Author bios present and linked to verifiable external profiles
• Content cites named sources throughout
• Brand mentioned in third-party editorial coverage, forums, or review platforms
• Earned media strategy in place (Reddit, LinkedIn, industry publications)

Scoring and tracking

• Audit scored across five dimensions (citation rate, content structure, technical access, authority, competitive position)
• Baseline score recorded with date
• Priority fix list created using an Impact vs. Effort matrix
• Quarterly full audit scheduled
• Monthly prompt monitoring scheduled (10 to 15 priority prompts)
• Triggered re-audit planned for any major site migration or URL change

This content was generated with the help of AI and it may contain mistakes