SEO and GEO audit: what a complete service must include

Key takeaways

• A technical SEO audit checks whether search engines and AI assistants can access, read and understand your site, and ranks every finding by impact.
• In 2026 a complete audit must review access agent by agent: GPTBot, OAI-SearchBot, ClaudeBot and PerplexityBot follow different rules from Googlebot.
• The cost of an audit depends on site size, report depth and whether fixes are included, not on a fixed rate.
• No serious audit promises citations in ChatGPT or better rankings. What it does is remove the technical barriers that make both impossible.

You asked for a quote for an SEO audit and received a PDF of PageSpeed Insights screenshots, a couple of Core Web Vitals charts and a list of errors in no order of priority? That is not an audit. Generic reports rarely answer the question that also matters today: whether the crawlers behind AI assistants such as ChatGPT, Perplexity or Claude can read your site the way Google does.

This guide sets out what a technical SEO audit with a GEO (Generative Engine Optimization) side must include: the steps, the tools, what it costs and what the final report has to contain.

What is a technical SEO audit?

A technical SEO audit is a systematic review of everything that decides whether a search engine or an AI assistant can access your site, understand it and treat it as a reliable source. The output is a diagnosis ordered by impact, not a list of “things that are wrong”.

In 2026 an SEO audit has two sides that share the same technical infrastructure:

• Classic SEO: crawling, indexing, architecture, performance and content for Google and Bing.
• GEO: whether AI crawlers can reach and read that same content, so that ChatGPT, Perplexity, Claude or Google’s AI Overviews can use it in an answer.

What is included in a classic SEO audit (technical and on-page)

The classic side is still the basis of any serious audit, and it has changed less than some headlines suggest:

• Crawling and indexing: which pages Google crawls, which a misconfigured robots.txt blocks by accident, and which fall out of the index because of a noindex directive or an incorrect canonical.
• Architecture: click depth, internal linking and URL hierarchy.
• Performance: Core Web Vitals, load time and resource weight.
• On-page SEO audit: search intent coverage, cannibalisation between pages, heading hierarchy and E‑E‑A‑T signals.

What a GEO audit adds

A GEO audit does not replace SEO; it extends it to systems that read without a browser. It checks three things:

1. Whether AI crawlers, both those that train models and those that answer in real time, are allowed in by robots.txt and by the firewall.
2. Whether the main content is in the raw HTML, because most AI crawlers do not execute JavaScript.
3. Whether structured data in JSON-LD states clearly what each page is and which entity stands behind it.

A site can have flawless technical SEO and still reach an AI crawler as an almost empty page, if the content is only painted on screen by client-side JavaScript.

SEO audit vs GEO audit: what each one checks

The steps of a complete SEO and GEO audit

A complete process usually follows this order, because each phase conditions the next: if a crawler cannot get in, it makes no difference how well optimised the content is.

1. Crawling and indexing: checking robots.txt, the sitemap, status codes and real coverage in the index.
2. Access per agent: robots.txt and firewall rules for each AI crawler, not only Googlebot.
3. Readability without JavaScript: what remains of each template once scripts are switched off.
4. Structured data: presence, validity and the schema.org types applied.
5. Performance: Core Web Vitals and server response time (TTFB).
6. Content and E‑E‑A‑T: search intent coverage, cannibalisation, authorship and answer-first structure.
7. Final report: findings prioritised by impact, with fixes and owners.

Steps 2 to 4 decide whether an AI crawler can get in, whether it can read the content and whether the structure helps or hinders that reading. They work as a knock-out filter: if something blocks access, no improvement in readability or structure can make up for it. That is why crawl rules must be reviewed agent by agent, as Google’s robots.txt documentation explains, rather than generically.

Which AI crawlers to review besides Googlebot

Not all AI bots do the same thing, and that difference matters when deciding what to block. Some train models, others answer questions in real time citing your page, and blocking the first guarantees nothing about the second:

A serious audit reviews robots.txt and firewall rules agent by agent, following the list of verified bots maintained by Cloudflare, and explains what blocking each category implies, without promising that stopping training crawls will keep that same assistant from citing you in an answer.

Which SEO audit tools are used (and which are not enough on their own)

A complete audit combines several categories of tools. Each one shows part of the problem and misses another:

None of these tools replaces the others, and none replaces a human reading of the result. This is, in fact, the most common problem in the sector: many reports sold as a “complete audit” are, in practice, a PageSpeed Insights screenshot with the agency’s logo on top, with no per-agent access analysis or structured data work behind it.

What AI can automate in an SEO audit (and what it cannot)

AI can greatly speed up crawling, cross-referencing data and drafting fixes (for example, generating corrected JSON-LD from faulty markup). What it cannot do on its own is decide which problem is most urgent for your particular business, or verify that an automatic fix has not introduced a new error.

An audit report generated without human review can sound convincing and be wrong, which is exactly the kind of missing verification that a GEO audit tries to detect on the client’s site.

The benefits of treating SEO and GEO as a single audit

Auditing classic SEO and the GEO side separately usually duplicates work: two crawls, two reports and two sets of priorities that do not always agree. A single technical process allows one coherent prioritisation and avoids contradictory recommendations between teams, for example the SEO team asking to open up a resource that the security team has just blocked for all bots indiscriminately.

The benefit is consistency and better use of the technical team’s time, not a promise of more traffic or better rankings.

How much does a technical SEO audit cost?

The cost of an SEO audit depends on three specific variables, not on a fixed rate:

Free SEO audit vs full audit

A free online SEO audit, run with an automated diagnostic tool, tells you in seconds whether AI crawlers can get in and read your pages. A full audit adds log analysis, content review and a prioritised plan. Before asking for a quote, it makes sense to start with the free diagnostic to pin down the real scope of your case.

What an SEO audit report must include

A useful report is not a flat list of errors. It must include:

If you want to understand first how all this fits into the GEO approach, read what GEO is and how it differs from traditional SEO. And if your question is more about index coverage, it is worth reviewing how to check index coverage in Search Console before commissioning a complete audit.

SEO audit checklist: what to check before you hire

Before accepting any SEO and GEO audit proposal, check that the report or the proposal:

• Explicitly mentions reviewing robots.txt per AI agent, not just for Googlebot.
• Tests firewall and anti-bot rules with AI user-agents.
• Checks the content with JavaScript switched off.
• Includes real validation of structured data, not just its presence.
• Orders priorities by impact, with owners, rather than presenting a flat list of errors.
• Justifies the price with specific variables (size, depth, scope), not a fixed figure with no explanation.
• Does not promise citation in AI assistants or ranking improvements as a direct result.

Frequently asked questions about SEO audits