AEO Audit for Agencies: How to Use It to Win More Clients

An AEO audit for agencies is a structured review of how well a prospect's pages and brand show up in AI answers from ChatGPT, Perplexity, Google AI Overviews and similar engines. Agencies use it to show a gap the prospect cannot see in their current reports, then turn the findings into a paid roadmap and monthly tracking.

Key takeaways

• An AEO audit has two layers: page readiness (can AI engines read and trust the page?) and brand presence (do they name and cite the brand?).
• It wins clients because it shows something new: competitors named in AI answers where the prospect is missing.
• Base every finding on evidence you can show: crawler rules, page structure, real answers and cited URLs.
• Do not promise ChatGPT rankings. AI answers vary between runs, so promise process and measurement instead.
• Turn the audit into a retainer with a baseline, a 90-day plan and monthly reports.

What is an AEO audit?

An AEO audit checks whether AI answer engines can find, understand, trust and cite a website, and whether they already mention the brand. It covers both the pages and the brand's presence in answers.

Answer engine optimization (AEO) is the work of making content easy for AI tools to use as a direct answer. A close cousin is generative engine optimization (GEO), which Wikipedia defines as structuring content and managing online presence to improve visibility in responses from generative AI systems. Our glossary covers AEO in more depth.

An AEO audit usually has two parts:

1. Page-level audit. Can AI crawlers reach the page? Does it answer questions directly? Is it well structured, sourced and trustworthy?
2. Brand-level audit. When buyers ask AI engines real questions, is the brand named? Which sites are cited? How do rivals compare?

Why do AEO audits help agencies win clients?

AEO audits help because they reveal a problem that prospects' current dashboards do not show: most marketing teams can see rankings and clicks, but few can see which brands AI engines recommend.

Data on clicks makes the case. Pew Research Center tracked 900 U.S. adults' Google searches in March 2025. In its 2025 study, users clicked a traditional result on 8% of visits when an AI summary appeared, versus 15% without one. They clicked a link inside the summary on just 1% of visits. A prospect with strong rankings can still lose attention to the answer above them.

First-party tools only show part of the picture. Google's Search Console has a generative AI performance report with impressions in AI Overviews and AI Mode. But it does not show competitors or other engines such as ChatGPT. That gap is where an agency's audit adds value.

An audit also helps in the sale itself:

• It gives the first meeting a concrete agenda built on the prospect's own data.
• It turns a vague topic, "AI search", into a short list of fixes.
• It sets a baseline, so later reports can show real change.

What should an AEO audit include?

A strong AEO audit checks access, structure, evidence, trust and brand presence. The checklist below covers the core items and where each one comes from.

Grounding the checklist in sources

Each check above ties to a public source. OpenAI's crawler documentation explains OAI-SearchBot. Google's AI features documentation says a page must be "indexed and eligible to be shown in Google Search with a snippet" to be a supporting link. The GEO paper tested nine content changes, and adding quotations, statistics and cited sources did best, with visibility gains of up to 40% overall (Aggarwal et al., arXiv:2311.09735).

Google's helpful content guidance covers trust:

"Of these aspects, trust is most important. The others contribute to trust, but content doesn't necessarily have to demonstrate all of them."

E-E-A-T stands for experience, expertise, authoritativeness and trustworthiness, the qualities Google describes in its guidance on helpful content.

For background you can share with clients, Google Search Central covers AI features in Search and what they mean for websites in this official video.

How do you run an AEO audit step by step?

Run the audit in two passes, a fast page-level check for the first call followed by a deeper brand study over several weeks, which keeps the sales cycle moving without compromising rigour.

1. Agree the scope

Pick 5 to 10 key pages, such as the home page, top product pages and best-performing guides. Agree on the markets and languages that matter. Ask which competitors the client loses deals to and which questions buyers raise on sales calls, because those become your competitor list and your first prompts.

2. Audit each page

Score each page against the checklist, documenting crawler rules, heading structure, direct answers, citations, structured data and trust signals, and record the exact remediation for each issue.

3. Build a buyer prompt set

Write 30 to 50 questions a real buyer might ask AI assistants, including "best", comparison, "how to" and "alternatives to" prompts, and keep most of them unbranded.

4. Track answers over time

Run each prompt daily for two to four weeks on the engines the client's buyers use. Because generated answers vary, a single run is statistically insufficient. OpenAI's own API guide says its model outputs "may differ from request to request" by default.

5. Add competitors and analyse

Track three to five competitors on the identical prompts, compare mention rate, citation rate, position and share of voice, and list the third-party domains the engines cite most frequently.

6. Write the roadmap

Group recommendations into quick wins, content development and off-site authority building, prioritizing the three largest gaps.

How do you present AEO audit findings in a pitch?

Lead with the competitive gap, then present the remediation, because "your rival is named and you are not" is easier for a prospect to grasp than a list of technical issues.

A clear audit report often follows this order:

1. One-page summary. The main gap, the score and the three priority fixes.
2. Competitive view. Share of voice and mention rate against rivals, per engine.
3. Real answers. Two or three example answers where rivals appear and the client does not.
4. Page findings. The page-level issues, ranked by effort and impact.
5. 90-day plan. What you will fix, what you will track and how you will report.

Be honest about limits. Explain that AI answers change between runs, and that you report rates over many runs. A 2024 study found accuracy swings of up to 15% across repeated runs of LLMs set to be deterministic (Atil et al., arXiv:2408.04667). Saying this up front builds trust and heads off questions about a single screenshot.

What does a good AEO audit finding look like?

A good finding names the problem, presents the evidence and recommends one specific fix, and it should make sense to a CMO within seconds. The example below is hypothetical, written to illustrate the format rather than drawn from client data.

Weak finding

"Your content is not optimized for AI search." This is vague, giving the client nothing to verify and nothing to implement.

Strong finding

"For the prompt 'best payroll tool for a small team', the brand was named in 3 of 20 answers. Two rivals were named in more than half. The cited pages were two review sites and a rival's guide. Your payroll page does not answer the question in its first lines, and it cites no sources. Fix: add a direct answer at the top, a comparison table and two cited facts, then pitch the two review sites."

Why the strong version works

• It uses the prospect's own buyer question as the evidence base.
• It quantifies the competitive gap in concrete numbers.
• It identifies the exact pages and sources behind the answer.
• It ends with a specific fix the team can implement this month.

Write every finding this way, then prioritize by effort and impact, placing two or three quick wins at the top so the client sees early progress.

How do you package and price AEO audits?

Package the audit as a clear entry point to ongoing work; because we have no sourced data on what agencies charge, set prices from your own delivery costs and market positioning.

Common packaging models include:

• Free snapshot. One or two pages scored, plus one competitor gap. Used to earn a meeting.
• Paid full audit. Page audit, prompt tracking over several weeks, competitor study and roadmap.
• Audit plus retainer. The audit becomes month one of an ongoing service with fixes and reports.

Whatever you choose, tie the price to effort, since brand-level tracking requires weeks of data collection and therefore costs more to deliver than a page check.

What mistakes should agencies avoid with AEO audits?

The biggest mistakes are over-promising, relying on a single answer and selling unproven tactics, each of which can undermine trust with a new client.

• Promising positions. No AI engine publishes how it picks brands. Promise a process and measured change, not "number one in ChatGPT".
• Using one screenshot as proof. One answer is one sample. Show rates across many runs.
• Selling llms.txt as a Google fix. Google's AI optimization guide says Google Search does not use llms.txt files, and they neither help nor harm visibility in Google Search.
• Keyword stuffing. The GEO paper found traditional tactics such as keyword stuffing offered little to no improvement in generative engine answers.
• Ignoring SEO basics. Google says its AI features are rooted in its core Search ranking and quality systems.

How do you turn an audit into a retainer?

Turn the audit baseline into monthly tracking and reporting: the audit establishes where the client stands, and the retainer demonstrates change over time.

Set up the client's prompt set, rivals and markets once. Then report each month on mention rate, citation rate, share of voice and sentiment against the baseline. Re-audit key pages after you change them, and show the score move. Tie the work to results the client cares about, such as visits from AI assistants and leads.

How can agencies run AEO audits with AnswerixAI?

AnswerixAI covers both layers of the audit. Our team built Site Audit to score any public URL from 0 to 100 on 47 AEO and GEO signals in five categories: Structure (25%), Content (25%), Authority (20%), E-E-A-T (20%) and Trust (10%). Each failing signal comes with a fix. When we built the rubric, we grouped the signals into those five weighted categories so a client can see at a glance where a page is weak. You can audit a competitor's URL on the same rubric and compare the two reports.

For the brand layer, each client is a brand in your workspace with its own prompts, competitors and market. Prompts are re-run daily. Growth and Business track ChatGPT, Perplexity, Microsoft Copilot, Google AI Overviews, Google AI Mode, Gemini and Grok. Starter tracks ChatGPT and Perplexity.

Reporting is built for clients. You can generate a report from four templates: Weekly Visibility Summary, Monthly Executive Report, Competitor Benchmark and Citation & Sources Report. Each downloads as a PDF. Growth and Business add white-label reports. Every plan includes unlimited seats and a 14-day free trial. See how it fits agency work on our agencies page, or compare plans on pricing.

Before your first client audit, read our guide to AI visibility metrics so your reports use clear, defensible numbers.