Because the score counts requirements rather than issues, some intuitive ideas about it are wrong. Knowing which saves wasted effort.
Things that move it
- Clearing every failure of a criterion. That criterion becomes a pass, and your score rises by roughly two points.
- A new kind of failure. Break a requirement that was passing, anywhere in the pages scanned, and you lose roughly two points.
- Which pages you scan. A page with a video, a form or a data table brings more criteria into play and more chances to fail them.
Things that do not
- Fixing some instances but not all. Fixing forty-nine of fifty unlabelled fields leaves the criterion failed. The score does not move until the last one is fixed. Your instance counts will still drop, which is real progress, but the score will not show it.
- How serious an issue is. Impact levels help you decide what matters most to people. They do not weight the score. A minor failure and a critical failure of different criteria cost the same.
- How many instances there are. A criterion failed once and a criterion failed a thousand times cost the same.
- Adding more pages that use the same templates. If a template fails a criterion, it has already failed.
Planning work that raises the score
- Open your expert audit and list every failing criterion.
- For each one, count how many instances are left. Criteria with only a few remaining instances are the quickest points available.
- For each one, find where it comes from. A criterion failing only in a shared template is usually one fix.
- Then work by impact level, because the score is not the only thing that matters.
Do not let the score set every priority
A critical barrier on your checkout that stops blind visitors paying matters more than a minor failure that would add the same two points. Use the score to track progress, and impact levels and your visitors to decide what to fix first.