Tuesday, October 12, 2010

IE Below 50%, But Not Universally

Internet Explorer logo

Perhaps you've seen the news, read the tweets, heard web developers shouting from rooftops (which is a heck of climb from the caves in which they are usually kept) — Microsoft Internet Explorer, the browser that has caused developers so much strife, has dropped below 50% market share. It's looking like 1998 all over again.

Catch up on the headlines:

There has been some good that has come out of this. Internet Explorer 9 (beta) has been pushing hard to speed up its scripting engine and better support existing and emerging web standards. This has been primarily due to the competition (and shaming) from Chrome, Safari and Opera. As we see the features, speed, and even interfaces of each new generation of browser start to match one another, developers and users will have a far more level playing field. This is assuming, of course, that trend continues and we don't see the Balkanization of the web again.

Despite all this, I am writing this post as a caution to developers. Already, many have been commenting on how great it will be to dump support for Internet Explorer altogether. Many developers seem to think that we are at that point now. Except we aren't. 50% isn't some magic number to drop support, especially not when developers often build features for the browsers that have far less market share (Safari, Chrome).

This blog only sees 14% of its traffic come from Internet Explorer, which doesn't surprise me given the subject matter. My professional site, which is at the primary domain used by this blog, has double that at 29%. In a vacuum, the much-publicized numbers seem quite real. However, as I review the logs of client sites, I can see a pretty simple trend. Clients whose primary business is B2B are trending between 60% and 80%, with some outliers in the 90% range, for Internet Explorer use. B2C sites enjoy lower IE numbers, hovering around 50-60%.

Remember that the B2B sites are probably visited by users in closed corporate environments where Internet Explorer is mandated. For many very valid reasons, IE has just stuck in that world and will probably continue to do so. Those same users may very well use another browser at home. We need to be careful that we are not excluding those users who either have no choice but to use IE or who aren't techs that can just swap browsers as it suits them.

To be clear: Don't drop your support for Internet Explorer. Make sure you still include it in your testing plan and have provisions for users who visit with IE — even if the experience isn't ideal.

Related

No comments:

Post a Comment