Skip to content
RU
WCAG 2.2

Somebody could not use your form yesterday

They did not write in. Accessibility audits the pages you choose against WCAG 2.2 on a schedule and tells you which barriers a machine can prove are there — with the element count, the success criterion, and what changed since the last run.

Three pages on the free plan. No card required.

One barrier, not one per element

Thirty-three images without alt text is one problem on your list, with thirty-three next to it. Filing it thirty-three times would bury the other five — and would make getting it down to five look like nothing had happened.

Undecided is its own answer

Contrast over a background image cannot be computed. That is neither a pass nor a failure, so it is listed apart, marked for a human, and left out of the score. Counting it either way would be a guess wearing a number.

A machine finds a minority of it

Automated rules cover the machine-checkable part of WCAG. A page can pass every one of them and still be unusable with a keyboard or a screen reader, so this reports what it can prove and never says your site is accessible. That sentence needs a person.

How it works

Point it at the pages people actually use

The form, the checkout, the sign-up. A barrier on a page nobody reaches costs nobody anything; a barrier on the one page that takes money costs you every time.

The page is rendered, not just fetched

A real browser with your stylesheets and images loaded, because contrast is the most common failure there is and it cannot be measured on unstyled markup. Checking it without CSS would not be a cheaper answer, it would be a wrong one.

A failed run is reported as a failed run

If the audit could not run — the page timed out, a bot wall answered — nothing is closed, nothing is scored and the pass says so. A tool that turns its own failure into a clean bill of health is worse than no tool.

You are told what changed, not just what is

A barrier that was fixed closes. One that comes back after a release is marked as returned rather than as new — a regression is worse news than a first sighting, and it should not read the same.

What it does not do

  • It does not certify anything and does not make you compliant with any law. It reports machine-checkable barriers on the pages you chose.
  • It does not replace testing with a keyboard and a screen reader, and it does not pretend the two overlap more than they do.
  • It does not fix your pages. It tells you which rule failed, on how many elements, and links the criterion.