An AI SEO audit checks whether AI search engines mention, cite, or recommend your brand when people ask relevant questions. The audit takes 30-60 minutes, costs nothing, and produces a baseline you can measure all future AI SEO work against. You query ChatGPT, Gemini, and Perplexity with the questions your customers ask, document what comes back, and identify the gaps between your current visibility and where you need to be. Most brands that run their first audit discover they are either completely invisible or mentioned but never recommended.
This guide walks through the audit step by step, explains what each result means, and shows you how to turn audit findings into an action plan.
Step 1: Choose Your Audit Queries
Start with 10-15 queries that represent how your customers search. These should mirror what a buyer would ask an AI search engine, not the keywords you track in Google.
Category queries: "Best [your category] tools in 2026," "top [your category] platforms." These reveal whether your brand appears in category-level recommendations.
Comparison queries: "[Your brand] vs [competitor]," "[competitor] alternatives." These reveal how AI search engines position you relative to competitors.
Problem queries: "How to [problem you solve]," "why is [pain point] hard." These reveal whether your brand appears when buyers describe their problem rather than searching for a solution by name.
Intent-specific queries: "Best [your category] for [specific use case]," "[your category] for small business." These reveal whether AI search engines match your brand to specific buyer intents or only mention you generically.
What to do: Write out your 10-15 queries before opening any AI search engine. Having the list ready prevents ad-hoc searching that misses important angles.
Step 2: Query Each AI Search Engine
Run each query through at least three AI search engines. Each engine has different retrieval behaviors and source preferences, so results on one engine do not predict results on others.
For each query on each engine, record:
- Does your brand appear? Yes or no. If no, note which competitors do appear
- In what position? First recommendation, second, third, or listed without endorsement
- With what sentiment? Positive ("strong option for..."), neutral ("one of several..."), or negative ("limited in...")
- What sources are cited? Note the URLs the engine links to. These reveal which third-party sources are driving recommendations
- Is your brand cited or just mentioned? A citation means the engine links to your website. A mention means your brand name appears but without a link
Run each query twice on different days. AI search results are nondeterministic, meaning the same query can produce different results on different runs. Two data points are better than one.
Step 3: Analyze Your Results
After completing the queries, you will have one of four results for each query on each engine. Each result points to a different type of gap.
Invisible (not mentioned at all). AI search engines did not find your brand for this query. This is a discoverability problem: either your content does not rank on Google for the sub-queries AI generates, or you lack presence on the third-party sources AI search engines check. Fix this by improving SEO for these queries and building third-party presence on review sites, Reddit, and editorial sources.
Mentioned but not recommended. AI search engines found your brand but did not position you as a solution. This is an intent-match problem: your content talks about your product generically but does not connect features to the specific use case the query describes. Fix this by creating content that addresses specific buyer intents, not just product features.
Recommended but not cited. AI search engines recommend your brand but do not link to your website. This means the recommendation came entirely from third-party sources. This is fine for visibility but means you are not capturing the referral traffic. Fix this by ensuring your website has pages that directly answer these queries so AI search engines have a reason to link to you.
Recommended and cited. This is the target state. AI search engines recommend your brand and link to your website. Maintain this by keeping the cited content fresh (updated within the last 30 days) and continuing to build third-party presence.
Step 4: Map Your Competitor Landscape
For every query where a competitor appears and you do not, document which competitors show up and what sources AI search engines cite when recommending them. This reveals two things: which brands you are competing against in AI search (they may be different from your Google competitors), and which third-party sources you need to be on.
Common patterns:
- Competitor has G2/Capterra reviews, you do not. AI search engines check review sites during the recommendation stage. Build your review presence.
- Competitor is mentioned on Reddit, you are not. Reddit is the most-cited domain for ChatGPT. Build authentic Reddit presence.
- Competitor has comparison content, you do not. Comparison pages that cover the full competitive landscape earn more AI citations than product pages. Publish your own comparison content.
- Competitor has recent content, yours is outdated. AI search engines prefer content published within the last 30 days. Update your key pages.
Step 5: Build Your Action Plan
Prioritize actions based on audit findings, starting with the gaps that affect the most queries.
If you are invisible on most queries: Focus on SEO fundamentals first (indexing, domain authority, content quality), then build third-party presence (reviews, Reddit, editorial). You need to be discoverable before you can be recommended.
If you are mentioned but not recommended: Create intent-specific content that connects your product to specific use cases. Publish comparison content that positions you clearly for the queries buyers ask.
If you are visible on some engines but not others: Target the missing engines specifically. Each AI search engine has different source preferences. ChatGPT favors Reddit and brand websites. Gemini favors structured data. Perplexity favors news and YouTube. Build presence on the sources each missing engine trusts.
If competitors dominate and you are a secondary mention: Strengthen your position on the specific sources AI search engines cite for those queries. Get more reviews, publish more comparison content, and build Reddit presence in your category.
Step 6: Schedule Follow-Up Audits
An initial audit is a snapshot. AI search results change over time as you publish content, build third-party presence, and competitors do the same. Schedule follow-up audits to track progress.
Monthly mini-audits: Re-run your top 5 queries across three engines. Takes 15 minutes. Automated tracking makes this even faster by monitoring changes over time without manual querying.
Quarterly full audits: Re-run all 10-15 queries with the same methodology as your initial audit. Compare against the baseline. Document which gaps closed and which remain.
Loudmink is an AEO platform that automates this audit process across up to 5 AI search engines every 24 hours, tracking changes over time so you do not need to run manual checks. Run a free scan to see where you stand. Plans from $99/mo as of June 2026.
Frequently Asked Questions
How often should I run an AI SEO audit?
Run a full audit (10-15 queries, 3+ engines) quarterly. Run mini-audits (top 5 queries, 3 engines) monthly. If you are actively building AI search presence (publishing content, building Reddit presence, requesting reviews), monthly mini-audits help you track whether actions are producing results.
Can I use free tools for an AI SEO audit?
Yes. The manual approach described in this guide is entirely free. HubSpot AEO Grader and Amplitude AI Visibility also offer free basic checks as of June 2026. Paid AEO platforms add automated scheduling, historical tracking, and multi-engine coverage that manual checking cannot replicate at scale.
What if my brand appears on ChatGPT but not Perplexity?
This is common. AI search engines have different retrieval behaviors and source preferences. ChatGPT favors Reddit and brand websites. Perplexity favors news sources and YouTube. If you are visible on ChatGPT but not Perplexity, focus on building presence on the sources Perplexity trusts: editorial coverage, YouTube content, and very fresh website content (Perplexity applies the most aggressive time decay).
Should I audit my competitors too?
Yes. Auditing competitors reveals which brands AI search engines recommend instead of you, what sources drive those recommendations, and where the gaps in competitive coverage exist. Include your top 3-5 competitors in the same audit queries to build a complete picture.