Self-authored VPATs get rejected because reviewers can see the tells in under five minutes. The six most common: every row marked Supports, an empty remarks column, the wrong edition, an outdated template version, no statement of what was tested and how, and no named tester.

The reviewer's checklist

University accessibility offices and agency 508 coordinators read dozens of these documents a year. They do not start with the criteria. They start with the signals below, and a document that fails two of them is usually sent back before anyone reads row one.

1. Supports on every row

No real product meets every WCAG 2.x Level A and AA criterion. A column of Supports tells the reviewer that nobody tested, or that the author did not understand the criteria. Reviewers expect Partially Supports entries, and they read the remarks on those first.

2. An empty remarks column

The template requires an explanation for every result other than Supports. Blank remarks mean the author cannot say what fails or for whom. A useful remark names the screen, the control, what happens for a keyboard or screen reader user, and whether a fix is planned.

3. The wrong edition

A WCAG edition sent to a US federal buyer is missing the Section 508 chapters. A WCAG edition sent to an EU buyer has no EN 301 549 mapping. The reviewer cannot complete their checklist, so the document comes back. The editions are explained here.

4. An old template version

VPAT 2.5 is current at the time of writing. A document on 2.4 or earlier suggests it was copied from an old report or a competitor's file, and it may be missing criteria that later template versions added.

5. No tested-environment statement

The evaluation methods section should list browsers, operating systems, assistive technology and versions, and the dates of testing. "Internal review" is not a method. Without this, the reviewer has no way to reproduce a single result.

6. No named tester

A report that names the person or firm who did the testing carries accountability. A report attributed to "the product team" carries none, and the reviewer treats it as marketing.

What a reviewer does with a rejected VPAT

  • Asks the vendor for a third-party assessment, which restarts the clock.
  • Runs their own testing and compares it with the vendor's claims. Discrepancies are remembered at renewal.
  • In higher education, may require an alternative access plan for affected users before the purchase can proceed.
  • In competitive procurement, scores the vendor lower on accessibility than a competitor with an accurate report.

What passes

A current template. The right edition for the buyer. A realistic mix of Supports and Partially Supports, with a remark on every non-conforming row that a developer could act on. A tested-environment statement with dates and versions. A named, independent tester. And an evidence report behind it, so when the reviewer asks a follow-up question the answer already exists.