If you aren’t familiar with it, WordPress publishes a showcase featuring prominent sites that have been built using WordPress. The goal of the showcase, as stated in the submissions page, is to “show the world what can be done with WordPress and help demonstrate the tremendous capabilities it has as a publishing platform.”

This may shock you, but there are no accessibility criteria for selection. I know – I just shook your world.

I’ve just finished doing a *very* quick top-level assessment of all 100 sites in the showcase. They range from some really quite good sites, with strong accessibility, to…really, really not.

It is probably a good representation of what can be done with WordPress. The reality is that WordPress is extremely flexible, so as a showcase of possibilities, the gamut from straightforward and accessible to OMG WTF is pretty realistic.

If anybody wants to share their thoughts or get involved in figuring this out, the process is being tracked in a Github issue for the showcase.


The goal was to get through a lot of sites quickly. So I ran them through WebAIM’s WAVE, gave them a quick visual and keyboard inspection, and reviewed the errors thrown. If there were a lot of errors thrown, I generally skipped any additional steps.

I made comments on each individual site as I felt it was necessary – to explain why I didn’t mark a site with very few errors as a candidate for further testing or to indicate why a site with more errors was one I still considered worth looking at.

Some statistics

For obvious reasons, I didn’t do a full assessment of these sites. Who’s got time for that? So these numbers aren’t going to be truly representative of accessibility; just of what WebAIM’s WAVE tool found on the site.

The largest number of errors found was a whopping 189. Of those 189 errors, there were 93 missing alternative texts, 88 empty links, 2 empty headings, and 5 missing form labels. I guess that site focused on not adding content to their site.

The fewest was 0. Now, I should mention that in fact there were numerous sites that showed no errors in WAVE; but on most of those, this was because they were in fact unusable sites that only contained errors not detectable by WAVE.

Overall, I identified 22 of the 100 showcase sites as worth a closer look at. That doesn’t mean that these sites are accessible; just that they demonstrated some attention to accessibility, so it may be worthwhile to tag them as accessible in the directory.

The biggest disappointments

  1. So much animation. I do the majority of my browsing with prefers reduced animation disabled. If there’s anything that’s likely to make me leave your site, it’s undesired animations. But dozens of these sites featured continuous animations and videos that ignored my preferences and had no pause mechanisms. Let me be blunt: respecting user preferences is just not that hard, and while it does mean your site doesn’t behave the way you want it to, it only does that for people who need it. Get on it.
  2. Lack of visible focus. Again, dozens of sites that tested pretty well, but a 30 second keyboard check wiped them out, because I couldn’t tell where I was. Again: easy to do, only impacts people who need it.
  3. Inaccessible menus. The site menu is probably one of the most critical parts of your site. The number of sites where submenus couldn’t be reached or menu toggles were constructed without using any kind of control element was disappointing.
  4. Empty links and missing or low quality alt attributes. No surprises here, but there were a few sites that were actually quite accessible – except that every link was an image, and the alternative text was something like “cow image.”

I think what’s most disappointing is that I looked at all 100 sites in under 3 hours, and a significant percentage of the problems found – if I had access and permission to make needed changes – could be completed in 20 minutes. (That’s each, not total. Don’t be silly.)

And the vast majority of those changes would have no visual impact on the site for most users.

What did I learn?

The showcase is basically divided into three different groups of sites:

  1. Sites that have paid attention to accessibility. They may have a few issues here and there, because people are imperfect, but the overall experience is good.
  2. Sites that have not paid attention to accessibility. They’ve had pretty much exactly the same design and build attention as the first category, but nobody gave a thought to accessibility. There is no reason this group of sites couldn’t be accessible beyond this lack of attention to detail.
  3. Sites that are not intended to display information. There is a handful of sites that are so focused on unusual design, animation, and unexpected behaviors, that I don’t really see them as websites. They may superficially behave like a website, but the actual intention is to show off a design aesthetic or technological sophistication. But they aren’t a place you would go to get information. That doesn’t excuse the lack; but it is a different type of aim than a site that’s trying to build an information resource but is failing at accessibility.


The goal of this process was to start to identify sites that might be tagged as accessible in the showcase. From a practical standpoint, we’re probably going to have to set criteria that are well short of a full assessment. We’re not in the position to provide hundreds of hours of free consulting, after all.

One of the desired goals by some is to only allow accessible sites in the showcase. I’m not sure that’s actually realistic or desirable. Is the purpose of the showcase to say that these are great, or is it to say that these are things within the scope of what’s possible for WordPress? These are representative of what people really do; if we can highlight those that are more accessible, maybe we can encourage some of those who don’t meet those criteria to work for it.

But the labeling will be crucial. It would be irresponsible to label any site as “accessible” without thorough vetting; and it’s also mean-spirited to disqualify a site that’s really pretty good but not perfect. Finding a clear label and an appropriate set of qualifying criteria is important.