How much does an SEO audit cost, and what should be in it

That range is wide because "audit" describes at least four different products. Knowing which one you are being sold is most of the purchasing decision.

What actually drives the price

Four things move the number, and only one of them is site size.

Page count, but not linearly. A 40-page site and a 400-page site take similar effort to audit if the 400 pages come from one template. A 40-page site assembled by three agencies over six years takes longer than either. What costs money is inconsistency, not volume.

Number of languages and markets. This is the biggest multiplier and the one most often underestimated. Auditing a site in four languages is not four times the work, because the technical layer is shared. It is roughly double, because hreflang, canonical logic, per-market keyword research, and local competitor sets all have to be checked separately. Anyone quoting the same price for a single-market and a four-market audit has not looked at the site.

Whether backlinks are in scope. Link analysis needs paid tooling and manual review of referring domains. It is frequently sold separately, and frequently left out of a quote without anyone mentioning it. Ask.

Whether you get a plan or a list. A findings document costs less to produce than a prioritised roadmap with effort estimates and expected impact. The second one is what you can actually hand to a developer.

Rough European benchmarks

These are market observations rather than a price list. Treat them as orientation.

Prices in Germany and the UK sit at the upper end of each band. Southern and Eastern Europe run lower for the same scope. A German buyer searching for audit costs and finding a €300 offer is looking at something different, not a bargain.

What a real audit contains

If you take one thing from this article, take this list. It is what to check a proposal against.

Technical. Indexation and crawlability, robots and sitemap accuracy, canonical logic, redirect chains, HTTP status coverage, Core Web Vitals with real field data rather than a single lab score, mobile rendering, structured data validity, and for multilingual sites, hreflang cluster integrity. That last one breaks constantly and almost nobody checks it.

Semantics. The actual queries your market uses, clustered by intent, mapped against the pages you have. The output should show you which clusters have no page at all. That gap list is usually where the money is.

Content. Coverage depth against what ranks, cannibalisation between your own pages, thin and orphaned pages, and whether the pages that rank actually answer the query in the first screen.

Commercial and conversion. Whether pricing, proof, and contact paths exist on the pages that receive traffic. An audit that stops at technical health tells you why you are not ranking, not why the traffic you have is not converting.

Competitors. Not a list of who ranks above you. A breakdown of what they have that you do not: pages, depth, links, structured data, freshness.

A prioritised plan. Every finding with an effort estimate and an expected effect, sorted so you can start on Monday. Without this you have a document, not a decision.

What a cheap audit skips

The €300 audit is usually a Screaming Frog or Semrush export with a template around it. There is nothing wrong with the tools. The problem is that a crawl reports symptoms without weighting them.

A tool will tell you that 47 pages have duplicate meta descriptions. It will not tell you that 44 of them are pagination that does not matter, and the other 3 are your money pages competing with each other. Sorting those apart is the work you are actually paying for.

Cheap audits also tend to skip anything that requires manual judgement: intent mapping, competitor content analysis, hreflang verification across languages, and whether the commercial pages make a case a buyer would act on.

When an audit is not the right purchase

Worth saying, because agencies rarely do.

A brand-new site with no traffic. There is nothing to audit. What you need is keyword research and an architecture, which is a different and usually cheaper piece of work.

A site you are about to rebuild. Audit the new one. Auditing something you are replacing in two months is spending twice.

When you already know the answer. If three people have told you the site is slow and the content is thin, you do not need a fourth document. Fix the two things, then measure.

When there is no capacity to act. An audit with no developer time behind it is an expensive way to feel informed. If nobody can implement for six months, buy the audit in six months.

The exception is a migration. Auditing before a replatform or domain change is the cheapest insurance in this field, because migration mistakes are the one category of SEO damage that is both severe and entirely avoidable.

How long it takes

An express check is a few days. A full single-market audit is typically two to three weeks, most of which is analysis rather than crawling. Multilingual runs three to five weeks.

Anyone offering a comprehensive audit in 48 hours is running a tool. Anyone taking three months has not scoped it.

How to read the deliverable

When it arrives, four questions separate a useful audit from an expensive PDF.

1. Is every finding prioritised, with a stated reason for the priority?
2. Does each item have an effort estimate a developer would recognise?
3. Can you tell what to do first, without asking?
4. Are the expected effects stated as ranges with reasoning, rather than as promises?

If the document ranks issues as "high, medium, low" with no explanation of the ranking, that ranking was probably produced by the tool.

One more test. A good audit tells you which of its recommendations it is least sure about. Search behaviour is not fully observable, and honest analysts say so. A document with no uncertainty in it has usually had the uncertainty edited out.

What to ask before you pay

• Which of the six areas above are in scope, and which are excluded?
• Is backlink analysis included or quoted separately?
• Do I get a prioritised roadmap or a findings list?
• Will you check hreflang and per-market keyword sets for each language, or audit one and

extrapolate?

• Do I get a walkthrough call, or just the document?
• Who implements, and is that quoted separately?

The hreflang question is the one that reveals most about a multilingual proposal. It is tedious, it is easy to skip, and it breaks silently on the majority of multilingual sites we look at.

---

Frequently asked questions

See also