Measured accessibility improvements

Tom, a colleague and IA in our User Experience group in Boston, took the time today to run some Wired News pages through the Bobby accessibility evaluator. For comparison, he tested both the old and new WN designs, along with Cnet’s News.com. As I noted in one of my own tests about a month ago, the site is still not perfect from an accessibility perspective. But the tests and comparisons Tom reported back show the new design will help us make significant progress compared with the old design (and the design of our competitor). The report showed the following:

  • Priority 1 errors
    dropped from 28 (old design) to 6 (new design) instances. (News.com = 91)
  • Priority 2 errors
    dropped from 147 to 11 instances. (News.com = 187)
  • Priority 3 errors
    dropped from 48 to 22 instances. (News.com = 75)

The point of this data is not to pat our team on the back by reporting how much work we’ve done on WN accessibility with the new design, nor to compare WN to our competitor. My point is that we’ve achieved these results without even putting that much effort or thought into making WN more accessible. Obviously, the more I learn about the benefits of these considerations, the more care I take in following accessibility guidelines. But by keeping all design and formatting details within the CSS, and using strict XML-based rules of XHTML, we’ve managed to dramatically improve our Bobby accessibility evaluation. Of course, the elimination of almost all tables from the markup, and a little more attention paid to alt tags helped. But the real improvement came simply from adhering much more closely to basic Web standards.

Ideally, we’d all be making significant improvements to the accessibility of our sites. It just makes pure business sense as a rapidly growing percentage of disabled and wireless users are accessing our sites through alternative browsing applications.