Some people call it "designing for our future selves", in order to get the message across that one day we will all benefit from a Web that has been designed by people who think of more than the visible. But in fact, some of us could even benefit today from an accessible Web: for instance in a car being read the route to somewhere from a website, by a browser that reads webpages aloud.
The current web is a terrible mess for accessibility. Try turning off images in your browser next time your use it in order to appreciate what blind people have to deal with. If you build websites, do it once a week to keep yourself alert.
There is no reason for building websites like this: it is not difficult to do it right. But even well-meaning website builders ask: How can I justify the extra cost for such a small percentage of the public?
And the answer is: Google. Looking at my website logs, it is clear that at least half of the visitors find the sites via Google. And what Google sees is exactly what a blind person sees. Google is a blind user; a billionaire blind user, with millions of friends who listen to his every word. If a blind user can't see your site, neither can Google, and your site will suffer.
Here are some examples of things to avoid if you want Google to like you (although not all of these improve accessibility, they almost all improve the web):
- Sites with no searchable content (e.g. Flash)
- Text-as-graphics (although Google handles a number of formats, it doesn't OCR).
- Frames lacking <noframes> content
- Java/Javascript dependencies (such as text generated by Javascript)
- Obviously CGI-driven sites (search engines tend to avoid them)
- Session-ID'd URLs
- Audio content (which is not searchable)
- Changeable sites. (The longer you stay in one place, the better your Google score. Don't move content! Don't change your URL's!)
In other words: accessibility isn't a costly nice-to-have extra: it is a bottom-line affecting essential.
(With thanks to Karsten M Self)
© Copyright Steven Pemberton, Amsterdam, 2003. All rights reserved.
First published in ACM/Interactions, January 2003