The honest answer is that an accessibility audit costs what the site costs to test, and sites vary more than buyers expect. But the market does publish numbers, the drivers are knowable, and the difference between a useful audit and a cheap one is easy to describe. This is what to expect, and what to ask.
1. What the market publishes #
Toronto agencies that publish prices put a website audit for a typical business site in the low thousands: Zwebra cites a range of roughly $2,400 to $6,000; PerfectPixels cites $2,000 to $8,000 for a site of ten to fifty pages. Those ranges are consistent with what we see in the market, and they describe the same thing: a manual audit of a small-to-medium marketing site.
Two things sit outside those ranges. An automated-only scan costs a fraction of that, or nothing, and answers a different question. A web application with authenticated journeys, or a public-sector site with hundreds of documents, is a larger engagement priced from its own scope.
2. What drives the price #
| Driver | Why it moves the price |
|---|---|
| Templates, not pages | A thousand pages on eight templates audit faster than fifty bespoke ones. Template count is the first question we ask. |
| Journeys | Enquiry, application, checkout, login, search: each journey is walked with a keyboard and a screen reader, end to end. Applications have many. |
| Documents | Every PDF, Word and slide file on the site is web content in law. Sampling them is quick; auditing an archive is a project of its own. |
| Assistive-technology depth | Screen readers on two platforms, zoom to 400%, high contrast, reduced motion, voice control. Each is a pass through the journeys. |
| The reader | A findings list for developers is quicker to write than a report a director can sign and a procurement office can file. Ask for both. |
| Re-testing | Verifying fixes after remediation is a second, smaller audit. Price it in from the start. |
3. What a cheap audit leaves out #
Automated tools detect about a third of WCAG failures, reliably: missing alternative text, contrast, structural problems. They cannot judge whether a link’s name is meaningful, whether a keyboard user can reach the menu, whether an error message helps, or whether a document reads in the right order. An audit that is a scan export with a logo on it answers the compliance-report question with a guess.
The tell is the findings. A real audit’s findings name the page, the element, the criterion, the assistive technology it was found with, and the fix. A scan’s findings name a rule and a count.
4. How we price one #
We quote a fixed fee after a short scoping call and a crawl of the site. The scope names the templates, the journeys and the document sample; the fee follows the scope, not the page count; and the price of re-testing after remediation is stated in the same quote. The audit page describes what is covered. We do not publish a rate card because a rate card would misprice most sites in one direction or the other, and because the scoping call usually makes the audit smaller than the buyer feared.
5. What remediation costs relative to the audit #
The audit is the small number. Remediation depends on what the audit finds and on whether the barriers live in templates, where one fix reaches every page, or in content, where they do not. A well-built site with a handful of template-level failures is often cheaper to fix than it was to audit. A site whose colour system fails contrast everywhere is a design decision, and we say so in the report rather than pricing it as a bug.
Two situations change the arithmetic. An audit before a redesign turns the findings into requirements, and accessibility built in costs a fraction of accessibility retrofitted. And an audit before a compliance-report deadline is what makes the report’s website question answerable; for an organization filing by 31 December 2026, it is the cheapest piece of evidence you will buy.
6. Questions to ask any auditor #
- Which WCAG version and level, and how do you report the criteria the regulation does not yet name?
- Which assistive technologies do you test with, on which journeys?
- Are documents in scope, and how many?
- Who reads the report, and is there a version for each reader?
- Is re-testing after remediation included, and at what price?
- Will the people who wrote the report fix the findings if we ask, and will they say which fixes are trivial?
If the answers are specific, the price is probably right for what you are getting. If they are not, the price is the least of the problems.