VPAT vs accessibility audit: what is the difference, and how they work together
Vendors often hear "VPAT" and "accessibility audit" used as if they were the same thing, and sellers sometimes blur them on purpose. They are different documents with different readers. This page explains each one plainly and shows how they fit together.
The difference in one line each
- An accessibility audit is an evaluation of your product against a standard, usually WCAG. Its output is a list of issues: where something fails, why, how severe it is, and how to fix it. Its reader is your team.
- A VPAT is ITI's template for reporting conformance. Once it is completed for a product, it becomes an Accessibility Conformance Report (ACR). It states, criterion by criterion, how well the product conforms. Its reader is your buyer.
An audit tells you what to fix. An ACR tells a customer where the product stands today.
Side by side
| Accessibility audit | VPAT / ACR | |
|---|---|---|
| Purpose | Find and explain barriers so they can be fixed | Disclose conformance to a buyer |
| Audience | Designers, engineers, product owners | Procurement, accessibility offices, legal |
| Format | No fixed format: issue lists, screenshots, code-level advice, severity | ITI VPAT 2.5Rev template (WCAG, 508, EU or INT edition) |
| Granularity | Individual defects, often many per criterion | One conformance level and remark per criterion |
| Shared publicly? | Rarely. It is internal working material. | Usually. Section508.gov advises making it easy to find on your website. |
| Who produces it | An auditor or your own trained team | The product owner, often written from an audit |
| Certification? | No. An audit is an opinion at a point in time. | No. ITI does not review, approve or certify VPATs. |
What an audit involves
A credible audit combines automated checks with manual testing: keyboard-only use, one or more screen readers, zoom and reflow, colour contrast, and a review of forms, error handling and dynamic content. W3C's Web Accessibility Initiative publishes a methodology for evaluating conformance, WCAG-EM (W3C WAI, WCAG-EM). It is also plain about tools: "Web accessibility evaluation tools can not determine accessibility, they can only assist in doing so" (W3C WAI, Selecting evaluation tools).
A good audit report gives each issue a location, the criterion it fails, its impact on users, and a recommended fix. That level of detail is exactly what a buyer does not need in an ACR. It also makes most audit reports unsuitable for sharing as they are.
What an ACR involves
An ACR condenses the evaluation into the template's structure. For each success criterion it gives one of four levels, Supports, Partially Supports, Does Not Support or Not Applicable, and a remark. It also states the product version, the report date and the evaluation methods (ITI). A good remark is short and checkable. It says what does not work, any workaround, and the plan.
A completed ACR is a summary statement made by the vendor, and ITI says the manufacturer is usually the best source for it. That is why the methods section matters. It is how a reader judges whether the statement rests on real testing. More on the format is in What is a VPAT?
How they work together
- Audit. Test the product version you are about to report on.
- Fix the quick wins. Many audit findings are small, and fixing them before the report improves it honestly.
- Write the ACR from the audit. Map each finding to its criterion, choose the conformance level, and summarise the issue in the remark.
- Publish the ACR. Keep the audit internal and use it to plan the remaining work.
- Repeat when the product changes. A new release needs re-testing in the areas that changed, then an updated ACR.
Skipping the audit gives you an ACR with nothing behind it. Reviewers notice remarks that are vague or rows that all say "Supports". Skipping the ACR leaves you with useful engineering notes and nothing a buyer can read.
"A buyer asked for a VPAT. Do I need an audit?"
You need an evaluation of some kind, because the report must say how you tested. Whether it has to be independent depends on the buyer. ITI notes that VPATs do not require third-party review, but a government or other solicitation may require a third-party audit. Follow the solicitation, and ask the issuer if it is unclear.
If your team has the skills, an in-house evaluation is legitimate, as long as the methods section is honest about it. If not, hire an auditor and have the ACR written from their findings. Published prices for both routes are in How to get a VPAT / ACR.
Red flags when buying either
- "VPAT certification" or a "certified VPAT". No such status exists.
- An "audit" that is only an automated scan.
- An ACR with no product version, no date, or no evaluation methods.
- A promise of "full compliance" before anyone has tested the product.
Where AccessRecord fits
AccessRecord is software for the ACR side: drafting it against the VPAT 2.5 criteria, publishing it at a stable public link, pinning it to the product version, and flagging it when a release moves past it. It is not an audit firm and it does not test your product. It works with your own testing or any auditor's findings. If you would rather hand the whole job over, there is a fixed-price option to have the report written, which includes a hands-on evaluation.
Sources
- ITI, VPAT 2.5Rev and FAQ
- W3C WAI, WCAG-EM overview and Selecting web accessibility evaluation tools
- Section508.gov, Accessibility Conformance Report