WebAIM - Web Accessibility In Mind

E-mail List Archives

Re: priority accessibility issues

for

From: Steve Green
Date: Apr 13, 2022 8:10AM


That's a question I get asked a lot, and I refuse to answer it because it is not a useful question. In particular, it is not what the person asking it really wants to know. And it hides a lot of complexity. So the answer is "why do you want to know?"



The person asking it usually wants to know something different, such as:



* In what sequence should we fix the bugs?

* Which bugs can we go live with and which must be fixed first?

* Can we leave some bugs permanently unfixed?



Once I get a sensible question like this, my answer is usually “if you are going to fix some or all of the bugs before going live, do so in the sequence that makes the most efficient use of development and testing resources.” You may choose to fix some low-impact bugs if they are in the same area of code as a high-impact bug you have decided to test.



Prioritisation of post-launch fixes is rather different and there are a lot of factors to consider. For a US website I might recommend prioritising all the issues (including false positives) that automated tools report in order to minimise the likelihood of a spurious legal claim. For each bug, you might want to assess which user group(s) are impacted, how large such groups are, the extent to which they are impacted, the importance of the inaccessible feature or content and whether there are workarounds. Most of that is subjective and some is unknowable. There is no right answer.



The important thing to explain to project managers (since it is usually they who ask the question) is that you can't just say that WCAG success criterion 1.3.1 is higher priority than 1.4.3 because it's level A rather than AA. Taking into account the criteria I mentioned above, some level AA non-conformances might be much more important than some level A ones on a particular website. And you certainly shouldn't generalise across websites.

Steve Green
Managing Director
Test Partners Ltd