Violations vs. manual-review items
A scan result has two different kinds of findings. They mean different things, and only one counts against your score.
Violations
These are specific failing elements the scanner detected with confidence — a button with no accessible name, an image missing alt text, insufficient contrast, and so on. They're grouped by WCAG rule, rated Critical / Serious / Moderate / Minor, and each comes with How to fix guidance. Violations are what your Open issues count and score reflect.
Work through these first: How to fix common issues.
Manual-review items
Some WCAG checks can't be decided by a machine — they need a human to look. The scan surfaces these separately, in a distinct (non-red) section headed:
Needs manual review — not counted as violations
with the note that these are *"checks to confirm by hand, not failures."*
A manual-review item is not a problem the scanner found. It's a prompt: *a person should verify this one.* These do not count toward your violation total or your score.
The honest bottom line
The product states this on every scan result, and it's worth repeating:
Automated detection of specific failing elements — not a full accessibility audit. Automated checks catch a portion of WCAG issues; a professional audit is still recommended for full conformance.
A clean automated scan is good, but it's not the same as full WCAG conformance. The manual-review items are exactly the parts automation can't judge — and a professional audit covers the rest.
Was this page helpful?
