Website Audit Cost: Prices, Scope and a Quote Checklist

For concrete reference points, the examples below include a free crawler tier, a £199 annual software license and a €197 fixed-price review advertised by its provider. They are published examples of different scopes, not an industry average or interchangeable offers. A bespoke audit needs a quote based on your site and the problem you want investigated.

What are you paying someone to investigate?

Begin with a question: “Why did important pages lose search impressions?”, “Where does the purchase flow fail?”, or “Can people use this form with a keyboard?” That question determines the evidence needed. A general homepage score may reveal useful clues, but it cannot establish that every page, account state or transaction works correctly.

Several of these scopes can overlap, but a supplier should say which ones are included. A five-page site can contain a difficult payment or account workflow; a large collection of similar static pages can be sampled by template. Page count matters alongside templates, integrations, languages, access states and the depth of investigation.

Published price examples: compare the unit and the scope

These public examples were checked on 3 November 2026. Keep the original currencies and confirm current taxes, limits, billing terms and the exact offer before paying. A tool license pays for use of the software; it does not purchase an expert’s analysis or implementation.

Sources: Chrome’s Lighthouse documentation, Screaming Frog pricing and GrowthFriction’s published offer. The service example illustrates how an inclusion/exclusion list affects the price comparison; its public listing does not independently establish the quality or suitability of its work for your project.

For custom work, request a fixed-scope project price or an hourly/day rate with a stated estimate and approval process for extra work. Ask for the same brief from each supplier. Do not infer that a freelancer, agency or low-priced automated report always includes the same categories or depth. A retainer can buy ongoing work, but it is not required for every agency audit.

What changes a professional audit quote?

• Coverage: How many representative templates, URLs, journeys, devices, languages and account states must be checked?
• Access and evidence: Is public browsing enough, or are search-console data, logs, analytics, staging accounts or stakeholder interviews needed?
• Investigation: Is the job an automated crawl, manual reproduction of a reported bug, a migration review or specialist testing?
• Deliverables: Will you receive raw exports, a prioritized explanation, developer-ready reproduction steps, a walkthrough or support answering questions?
• Implementation and verification: Are fixes and a later retest included, optional or separately quoted?
• Timing: When are access and inputs ready, what is the delivery date, and what does an urgent request change?

A focused diagnosis can be more appropriate than an all-purpose review when one problem is urgent. If Bing impressions dropped after a particular date, supply the affected pages and queries, comparable date ranges, release history and any indexing changes. Ask the reviewer to distinguish evidence from hypotheses rather than promise a penalty diagnosis or a return to previous rankings.

A worked budget: audit, fixes and retest

Here is a hypothetical time-based quote in US dollars. The rate and hours are assumptions for illustrating the math, not typical market rates or a supplier offer. Replace them with the actual estimates you receive.

For a fixed-fee quote, the same distinction still helps. Record the included work, assumptions, exclusions and change process. An attractive audit price can be only the first part of the project if every fix and follow-up is additional. Conversely, you may already have an internal developer who can implement an agreed plan. Count that time even if it is not a new invoice.

What free checks can and cannot establish

Free tools are useful for collecting evidence and narrowing a question. They are not all limited in the same way. Some support broad crawls or manual investigation; others inspect one page’s HTML. Choose the method for the question rather than assuming that free work catches a fixed percentage of every website’s problems.

Google’s PageSpeed Insights documentation distinguishes controlled lab data from real-user field data. Field data may be absent or fall back to origin-level data when a page has insufficient samples. Do not present a single lab score as proof of every visitor’s experience.

The current Grigora Website Grader labels its resource sample and estimated speed in the report. Use its findings as a starting point, then reproduce the relevant issue. A high score based on markup cannot replace checking the actual mobile layout, a completed form or the behavior of an authenticated application.

Accessibility needs a similarly clear boundary. W3C’s evaluation-tool guidance says tools cannot check every aspect automatically and can produce misleading results; human judgment is needed. Likewise, a public security-header or TLS check does not establish that an application has no vulnerabilities. Commission more specialized work when the problem and the site justify it, with agreed authorization and scope.

Ask for findings you can act on

Request a redacted example deliverable before buying, then check whether it would help your team make a change. A useful issue record should identify the affected page or template, explain the method and evidence, describe the consequence, suggest an actionable fix and state how to verify it. A screenshot and a number can be part of that record, but neither is a complete diagnosis.

• Priority: Why this issue should be addressed before another one, including confidence and relevant dependencies.
• Coverage: Which pages and states were checked, what was sampled, and what was not tested.
• Ownership: Who is expected to implement each fix and what access or input is needed.
• Verification: The acceptance check and whether a retest is included.
• Handoff: The report format, source exports where appropriate, a walkthrough and any follow-up allowance.

For example, “improve your SEO” is not a specific task. “These intended public articles return noindex in their response HTML; confirm the intended setting, change the template if incorrect, and inspect the rendered response again” is a testable finding. The action still depends on whether those pages were supposed to be indexed. Avoid applying every warning without understanding the site’s intent.

Send the same brief to each supplier

You can adapt this short request:

Please quote a review of [site and environment] to investigate [specific problem]. Include [templates/journeys/devices/languages], the evidence and access needed, deliverables, exclusions, delivery date, implementation options and retesting. State the currency, tax treatment, billing basis and process for extra work.

Include enough context to make the scope concrete: recent changes, the pages or journeys that matter, known symptoms and who can implement findings. For authenticated work, agree how access will be provided and removed and which environment may be tested. Keep confidential production information out of public audit submission forms.

Compare written answers line by line. If one quote samples three templates and another tests every critical journey across multiple states, their totals are not directly comparable. Resolve that difference before choosing solely on price.

Budget for action without assuming a return

An audit is valuable when the findings are relevant, correct and acted on. It can still identify no major problem, or lead to changes whose commercial effect is hard to isolate. Do not assume that more traffic produces proportionally more revenue or that an audit will pay for itself within a particular number of months.

Set an outcome appropriate to the issue: a broken flow works, intended pages become eligible for indexing, a measured bottleneck improves, or a documented accessibility barrier is removed. Keep a baseline and retest under comparable conditions. Where you measure business impact, distinguish observed changes from claims that the audit alone caused them. The mobile layout guide illustrates a practical reproduction-first approach, and the sitemap scope guide explains why intended indexing matters.

Website audit cost FAQs