Written by the AX4E accessibility team. Checked against the primary sources listed at the end of this page (W3C, U.S. Department of Justice, ISO and others). Last reviewed November 4, 2026. Read how we write and review guides in our editorial standards.
What a scanner does
An accessibility scanner loads your pages and checks the code against rules based on WCAG. It flags things it can detect with certainty: missing alternative text, form fields without labels, insufficient color contrast, empty links and buttons, missing page language, duplicate IDs and incorrect ARIA.
How much does it find?
More than people often assume, and less than vendors sometimes claim. Deque's analysis of nearly 300,000 issues from first-time audits found that automated testing identified about 57% of issues by volume. Other studies that count by WCAG success criterion rather than by issue volume put automated coverage lower. Either way, a meaningful share of barriers can only be found by a person.
What scans catch well
- Images with no alternative text at all
- Form fields with no label
- Text with insufficient color contrast
- Links and buttons with no accessible name
- Missing page language and title
- Broken ARIA references and invalid attributes
What needs a person
- Whether alternative text is accurate and useful
- Whether the heading structure and reading order make sense
- Whether everything works with a keyboard, with a visible focus indicator
- Whether custom widgets like menus, modals and carousels work with a screen reader
- Whether error messages are clear and instructions are understandable
- Whether captions are accurate and videos have the audio description they need
- Whether a whole task, like checking out or applying, can be completed
How to read a scan report
- Do not panic at the number. One template error repeated on 200 pages shows up as 200 errors and needs one fix.
- Group by cause. Sort issues by rule and by template or component, not by page.
- Prioritize by impact. Start with issues that block tasks: unlabeled form fields, keyboard traps, missing button names on key flows.
- Separate errors from alerts. Many tools flag "needs review" items that may or may not be problems.
- Re-scan after fixes to confirm and to catch regressions.
Scan vs audit
| Automated scan | Expert audit | |
|---|---|---|
| Speed | Minutes | Days to weeks |
| Coverage | Rule-detectable issues across many pages | All WCAG criteria on a representative sample |
| Assistive technology testing | No | Yes |
| Best for | Finding patterns, monitoring, regressions | Conformance claims, ACRs, remediation plans |
Make scanning a habit
Sites change daily. Scheduled scans catch new issues as content and code change, and adding checks to your development process stops them before they ship. Pair that with periodic manual reviews of your most important pages and flows.
Common questions
Can an automated scan tell me if my site is compliant?
No. A scan can find many issues and prove some are present, but it cannot confirm conformance, because many WCAG requirements need human judgment and assistive technology testing.
How often should we scan our website?
Scan at least monthly, and after major content or design changes. Larger or frequently updated sites benefit from continuous monitoring and checks in the development process.
Why does my scan show so many errors?
Repeated templates multiply counts. Group results by rule and component. A small number of template fixes often removes most of the reported errors.
Start with a free scan, finish with a person
Run a free automated scan of your homepage, then request a free assessment and we will review what the scan cannot see.
Run a free scan Request a free assessment