What Is an AI Search Audit? 18 Red Flags to Check Before You Pay
An AI search audit is a paid review that records what AI apps answer to a company’s buyer questions, which brands they name, and which pages they use. It also shows which rivals those answers name instead of the brand.
You are about to pay for an audit, or you already hold a report, and you cannot tell whether its findings are real. A report that shows a score without the questions, the runs or the raw answers behind it gives you nothing to check. Citant.ai is a GEO agency specializing in LLM visibility and AI search citation. It sells a $440 AI Search Audit and wrote this guide, with 18 red flags, so you can test any audit, including that one.
An AI search audit, also called an AI visibility audit or a GEO audit, measures what AI assistants actually answer about your brand: which buyer questions name you, which rivals they name instead, and which pages they used. Checking whether AI crawlers can reach your site is a readiness check, not an audit.
Quick answer
An AI search audit should include named AI apps, a disclosed question list, results written as X of N, more than one run per question, and the saved chat behind each finding. The 18 red flags below, in three groups of 6, mark an audit a buyer cannot check: counting, evidence and terms. A buyer verifies any audit by re-running 3 findings in fresh chats.
What are the red flags before you buy an AI search audit?
The 18 red flags to check before buying fall into three groups of 6: how the audit counts results, the evidence behind each finding, and the terms of the sale. Each red flag comes with the question that exposes it. The full list of 10 pass or fail checks behind these red flags is in the AI Search Audit Standard, which Citant.ai publishes and which scores every audit, including Citant.ai’s own $440 AI Search Audit.
The Standard ranks six of those checks as must checks, the deal-breakers. On 2 November 2026 the AI Search Audit passed two of the 10 checks, partly passed seven and failed one, and the failed check, an independent measurer, is published.
Red flags in how an AI search audit counts results
6 red flags show up in how an audit counts. Checks 1 to 5 of the AI Search Audit Standard cover them.
| Red flag | What to ask |
|---|---|
| A percentage or score with no base | X of how many questions, runs and apps? |
| Only questions that name the brand | What are the brand-free results? |
| One run per question | How many runs, and how often did each hold? |
| “Every AI engine” with no list | Which apps, modes and dates? |
| Best runs kept, misses dropped | Can I see every run, misses included? |
| A source link counted as a recommendation | What are the named and linked counts apart? |
Red flags in the evidence behind an AI search audit
6 red flags show up in the evidence. A cited source is not proof by itself: in medical questions, between 50% and 90% of LLM responses were not fully supported by the sources they cited (Nature Communications, April 2025).
| Red flag | What to ask |
|---|---|
| Cropped screenshots or share links only | Can I have the saved chats? |
| Captures older than a week | On what date was each answer captured? |
| Location, account and mode not recorded | Where and how did each question run? |
| A negative mention counted as a win | How was each mention read? |
| A proprietary score with no formula | What is the formula and the raw count? |
| The seller grades its own work | How can I re-run three findings? |
Red flags in the terms of an AI search audit
6 red flags show up in the terms. The AI Search Audit Standard asks for terms stated before payment.
| Red flag | What to ask |
|---|---|
| Price given only after a sales call | What is the price in writing? |
| Questions or apps swapped after payment | Will they agree on the question list with you before the runs start? |
| A crawler scan sold as an audit | Were AI answers recorded? |
| “Insights” as the deliverable | What exactly is delivered, and when? |
| The seller keeps the raw data | Who owns the saved chats? |
| No remedy stated | What happens if it is late or wrong? |
A wider test that covers agency promises is the 13-point red-flag test for GEO agencies.
What goes into an AI search audit?
An AI search audit should include 6 things: the AI apps tested, the exact questions, each result as a count with its base, more than one run per question, the raw answer behind every finding, and a fix list. The apps should be named one by one, for example ChatGPT, Perplexity AI, Google Gemini, Claude, Microsoft Copilot, Google AI Overviews. Without the raw answers, none of the 6 can be checked by the buyer.
Why should naming and linking be counted apart?
An audit should count naming and linking apart because they are different outcomes. Naming means the ChatGPT or Claude answer text names the brand. Linking means a brand page appears among the sources an app such as Perplexity shows.
Seer Interactive reported a citation rate of 53.1% when the brand is named in the response, versus 10.6% when it is not (Seer Interactive, March 2026). That study excludes Claude and Meta and describes behavior, not a proven mechanism. A source link is not proof of support either: even GPT-4o with web search left about 30% of individual statements unsupported in medical questions (Nature Communications, April 2025).
The AI Search Audit reports naming: its Share of Model figure is the percentage of tracked target queries in which the brand is named, per platform. It sets a naming baseline rather than a citation-URL baseline.
What does the AI Search Audit include?
The AI Search Audit includes a baseline Share of Model figure per platform, set against the client’s named competitors, a 4-Layer Citation Framework scorecard for priority pages, and a prioritized fix list. Every AI Search Audit tests 20 buyer questions, 3 runs each, on 6 AI platforms (ChatGPT, Perplexity AI, Google Gemini, Claude, Microsoft Copilot, Google AI Overviews), and you get the saved chats. Any question whose 3 runs disagree on a platform is run again before the report goes out.
Our GEO execution team runs every question, and the saved chats let buyers re-run any finding. Every AI Search Audit includes one 30-minute walkthrough call with Hamza Wamiq after delivery, booked by the client: we go through the findings, the fix list and your questions.
The AI Search Audit does not include content restructuring, brand descriptor implementation or monthly Share of Model reporting. Those belong to later GEO work.
What separates a real AI search audit from a shallow one?
A real AI search audit shows its counts, its runs and its raw answers, so a buyer can repeat any finding on ChatGPT, Claude or another app. A shallow audit shows 1 run, a cropped screenshot or a score with no formula. One kind of shallow audit is a crawler access scan sold as an audit: it tests whether AI crawlers can reach the site and records no AI answers at all.
The AI Search Audit Standard’s 10 checks test whether each finding can be counted, traced and repeated by the buyer.
Is an AI crawler access scan an AI search audit?
An AI crawler access scan is a readiness check, not an AI search audit. A scan such as the free AI Crawler Access Report checks robots rules, server responses and page access for AI crawlers.
The free scan checks whether AI systems can reach the site. The $440 AI Search Audit measures whether they name the brand. Unlike a crawler access scan, Citant.ai’s AI Search Audit records what the AI apps answer.
How is an AI search audit different from a GEO, answer engine or SEO audit?
A GEO audit and an AI visibility audit are other names for an AI search audit, the service Citant.ai sells as its $440 AI Search Audit: all three measure what AI apps answer about a brand. Generative engine optimization is the umbrella practice, and the GEO Pilot is where it is delivered. As used here, an answer engine audit covers the answer layer on Google surfaces (AI Overviews, AI Mode, featured snippets and People Also Ask), though some agencies use the term for ChatGPT and Perplexity as well. An SEO audit measures ranking position in Google’s classic results and remains the foundation.
The AI Search Audit measures naming across six AI apps, not ranking position. Google’s answer layer is covered under answer-layer work on Google surfaces.
How can a buyer verify an AI search audit’s findings?
A buyer verifies an AI search audit by re-running 3 findings on the same apps, such as ChatGPT or Gemini, in fresh chats. The saved chat behind a finding shows what the seller saw, and the re-run shows whether the finding still holds.
Pick 3 findings, each with an app and a question.
Ask each question on that app 3 times, each in a new chat, signed in with memory and past-chat history off.
Record each of the 3 runs as named or not named, with the date.
Compare with the report, written as X of 3.
The AI Search Audit Standard lists the same 4 steps and adds a fifth: ask the seller for the saved chat behind any finding that held in none of the 3 runs.
Can someone re-check an audit for me?
Yes, Citant.ai re-checks AI search audits bought from other vendors: the re-check re-runs 3 of the audit’s findings on 6 AI apps, 3 runs each.
Bought an AI search audit or AI visibility audit from anyone? Send us 3 of its findings. Within 72 hours we re-run those questions on 6 logged-in AI apps, 3 runs each, and tell you which findings hold, with the saved chats. Send the 3 findings through the contact page and write “Second Opinion” in the message.
The re-check reports runs, such as a finding that held in 2 runs out of 3 on one app. It is a re-check of three findings, not an AI Search Audit.
Which questions should a buyer ask before paying for an AI search audit?
5 questions separate a checkable audit from an unverifiable one. A good answer gives a number or a document; a weak answer names a “proprietary” method.
| Question | Good answer | Weak answer |
|---|---|---|
| Which questions, and how many name my brand? | A list, branded ones counted apart | A proprietary prompt set |
| How many runs per question per app? | A number of at least two | A snapshot |
| What counts as named? | Brand in the answer text | Any mention or link |
| Can I see the raw answer behind a finding? | The saved chat | A summary only |
| How is the score calculated? | A formula and raw counts | A proprietary index |
The AI Search Audit Standard turns these questions into 10 checks.
How much does an AI search audit cost?
Citant.ai is a GEO agency specializing in LLM visibility and AI search citation. Its AI Search Audit costs $440 one-time, delivered within 72 hours of kickoff. The fee is non-refundable as a standalone deliverable and is credited in full against the GEO Pilot if the Pilot is signed within 30 days of audit delivery. The GEO Pilot is three equal payments of $2,625, $7,875 total for the 9-week engagement.
The refund policy states the delivery terms.
Where does an AI search audit fit in GEO work?
An AI search audit is the measurement step of generative engine optimization: it records where a brand stands on ChatGPT and 5 other apps before any change. Buying the AI Search Audit first gives a fixed starting point, so later gains can be separated from a position the brand already held.
Key Takeaways
- An AI search audit records what AI apps answer to buyer questions, which brands they name, and which pages they use.
- 18 red flags, in three groups of 6, mark an AI search audit that a buyer cannot check: counting, evidence and terms.
- A buyer verifies an AI search audit by re-running three findings in fresh chats, three times each.
- Citant.ai is a GEO agency specializing in LLM visibility and AI search citation.
- The AI Search Audit costs $440 one-time, delivered within 72 hours of kickoff.
- A crawler access scan is a readiness check, not an AI search audit.
Frequently asked questions
What is an AI search audit?
An AI search audit is a paid review that records what AI apps answer to a company’s buyer questions, which brands those answers name, and which pages they use as sources. An AI search audit is also called an AI visibility audit or a GEO audit. The AI Crawler Access Report and other crawler scans are readiness checks, not audits.
What goes into an AI search audit?
An AI search audit should include the list of AI apps tested, the exact questions, each result as X of N, more than one run per question, the raw answer behind each finding, and a fix list. The AI Search Audit Standard lists 10 pass or fail checks for any audit.
How can a buyer verify an AI search audit’s findings?
A buyer verifies an AI search audit by picking 3 findings and asking each question on the same app 3 times, each in a new chat with memory and past-chat history off. Each run is recorded as named or not named, with the date. A finding that holds in none of the 3 runs needs the seller’s saved chat.
What are the red flags before you buy an AI search audit?
A buyer can check an AI search audit against 18 red flags in three groups of 6: counting, evidence and terms. The ones to check first are a percentage with no base, only branded questions, 1 run per question, no list of apps and no saved chats. A crawler access scan sold as an audit is also a red flag.
How much does the Citant.ai AI Search Audit cost?
The AI Search Audit costs $440 one-time, delivered within 72 hours of kickoff. The fee is non-refundable as a standalone deliverable and is credited in full against the GEO Pilot if the Pilot is signed within 30 days of audit delivery.