It really depends on the domain. That two percent can be the difference between wildly profitable and failing.
After all, Google still happily sells clicks to IE 6 users. If you're business relies on external ads and optimizing conversion rates, IE 6 remains a priority.
I seriously doubt IE6 support can be the difference between "wildly profitable and failing" for any business. Can you name one example of a failing company that started supporting IE6 and then became wildly profitable?
Sure. The example I was thinking of in particular was an internal project from about 5 years ago where their CSS2 layout completely blew up on IE6. However rather than get down-n-dirty to fix it, they demanded additional budget for a rewrite using tables.
But just last week, another subcontractor apparently didn't realize that our government client was stuck on Flash v9 and spent the next two weeks trying to sort it out.
And it hasn't happened in a few months, but we used to get occasional complaints from clients stuck on Safari v4.
Obviously it would be the other way around, discontinuing IE6 support might lead to falling. Think of random niche industries with ancient old hardware and no-one who cares about software updates.
Fair enough, but I still don't think a business that is wildly successful would fail by discontinuing IE6 support -- even in a "random niche industry" where IE6 users make up more than 2% of their customer base. In fact, I doubt any of those businesses are currently wild successes. But I guess that all depends on how you define success.
Google dropped IE6 support for YouTube over a year ago. They are dropping Google Apps support for IE7 on Aug 1st (that's today for some people reading this).
For search results it's probably more worthwhile and not as hard to support the IE's but for their other apps they say they are only supporting that two most recent browsers.
Google Search dropped mainstream support for IE6 last fall. IE6 users are served a degraded experience that is roughly on-par feature-wise with Google Search as it existed in May 2010, and new features are no longer rolled out to IE6.
And this is how IE6 will slowly be deprecated across the industry. Sure it's easy for an intern to get up on his soapbox and proclaim, "We should stop supporting IE6 because that browser came out while I was still in elementary school." But when you look at the how many people still continue to use it, deprecating support is on a case-by-case basis, and it is irresponsible as web developers to just ignore that it even exists.
I would actually argue that Google Search is still supporting IE6, just as a degraded experience. Dropping support all-together means they just ignore it exists, which is far from the truth and likely won't happen for 5+ years.
Well, under that standard, Google Search still supports Netscape 4.7, which is also served a (slightly different from IE6) degraded experience. I wouldn't be terribly surprised if it supports Netscape 1.0 and Lynx either.
After all, Google still happily sells clicks to IE 6 users. If you're business relies on external ads and optimizing conversion rates, IE 6 remains a priority.