We tend to agree easily that all accessibility issues are important, but we also acknowledge out of necessity that some issues are more critical than others. What we don’t always agree on is how to judge severity. There are very good reasons for that: it’s complicated! There’s no pretending that figuring out how important something is has an undercurrent of judging who matters more.

But that’s not really what we’re doing when we try to decide what issues are more important. It may feel that way, but it’s just not true. Judging severity shouldn’t have anything to do with who the user is or what their disability is.

I use a rubric for judging severity that’s based on three conditions: importance, ubiquity, and challenge.

Importance: Criticality of the content

Importance doesn’t measure the importance of the accessibility issue: it measures the importance of the content where the barrier exists. Do you have a color contrast error on a fluff tagline on the site, or do you have a color contrast error on the ‘Get Support’ link? The site tagline provides flavor and character, and it helps tell your audience who you are. But it’s not really critical content, and if somebody can’t see it, it’s unlikely to break their experience on your site.

Getting support, however, is likely to be a relatively critical service. When you need it, you need it, and there may be no substitute.

The content that’s blocked is more important, so the issue has higher severity.

Ubiquity: Frequency of occurrence

Ubiquity is about how common an issue is. I’ll stick with color contrast as an example, since the type of issue itself isn’t very relevant to this point. But color contrast is one where ubiquity comes into play a lot.

Where is the issue? Is it all of the body text on the site, or is it a few specific call out blocks? They’re all important, but it’s a big difference in scope.

A problem that somebody encounters constantly on your site is a bigger issue than what somebody encounters occasionally in a few side paths.

Challenge: Ability to Overcome

How hard is it for a user to get around this issue? There are relatively few accessibility issues that are actually impossible to work around. Most issues aren’t absolute barriers. But that’s kind of not the point: barriers are still present even if you can get past them.

Why do you have to climb over a wall while others walk through an open door?

But barriers are certainly not all equal. Color contrast is relatively easy to overcome: you can enlarge the fonts or switch your operating system into high contrast mode. Not everybody can do that, but there are lots of ways to do it.

Missing accessible names, on the other hand, may be almost impossible to get around. A fancy toggle with an ambiguous name so that you can’t really tell whether it’s checked or not? A button with no text that just leaves you guessing? Sure, you can just try ’em out… but that’s playing roulette, not working around a problem.

An issue that a user may simply be unable to overcome without assistance is extremely severe.

What about number of people blocked?

Yeah, that’s somewhat relevant. I know that a lot of people like numbers. Fix this issue, and 5% of users will no longer be blocked! But I’m deeply unconvinced by these numbers. Numbers surrounding the percentage of people with disabilities are not very reliable, and may have very little correlation to people coming to your website, anyway.

If I was going to look at a measurement that’s more focused on numbers, I’d pick “time to fix.” But I also can’t make those estimates for people. The technical challenges may not even be relevant. Fixing color contrast can be a 5 minute technical challenge, but a six week management debate.

Is there a conclusion here?

No. Issues are important in different ways, and every issue needs an independent assessment. I think it’s useful for people receiving my assessments to have an understanding of where a severity indicator comes from; but how they resolve those issues is often a more internal challenge, and not one that I can directly influence. As a consultant, it’s my job to identify issues, offer guidance on how to resolve them, and help the client strategize their solutions to the best of my ability. Providing some kind of indicator of severity is a key part of that last portion, but requires interpretation.