IE6: Dark corner of the hidden web

I was reviewing some trends on the web last week with regards to the utilization of Microsoft Internet Explorer (IE). The focus of the analysis was on how Internet Explorer 8 was be taken up  against previous versions such as IE7 and IE6. Note: I could not find any tracking data on IE5, which in itself is probably the answer to any trending questions.

From an application monitoring stand-point, IE is a rare example of "single use" or "single track" upgrade cycle. When I have performed application utilization tracking in the past, the primary difficulty was not in determining if an application had been used, but exactly what version. We used all sorted of tricks (EXE header capture as good example of what did not work) to determine the exact version of an application a user was actually using on a particular day/time/place/machine/environment.

Note: the "EXE header capture" process looked like a great idea at the time, unfortunately, a lot of application minor and sub-version information was NOT contained in the primary executable (sometimes it was just a stub) but rather in supporting DLL's and other binary resources.

With Internet Explorer, due to the integration with the OS, you can only use one version of Internet Explorer on a machine. Once you have upgraded to IE8, that's it. You can't use IE6 again - Unless you rebuild your machine. "There is no turning back - The only way is up!"

Or, you can always try Chrome.  Oops.

What did I say about a simple story? Now, we have to worry about multiple browsers.  Specifically, Firefox, Chrome, Safari and to a much lesser degree Opera.

Here is a sample of the two reports that I was shown;

These results were captured from the Global Stats site found here: http://www.w3counter.com/globalstats.php

And most interestingly for Internet Explorer 8 application compatibility story, here is a comparison of the different versions of Internet Explorer

Version 6 of Internet Explorer is definitely on the decline, with a month-on-month decline for the past three years. And, you will see (highlighted by the red circle) , as soon as IE8 was released, the biggest loser was IE6, with some  loses on IE7.

And, here is where I start to get into trouble. If you look at the source of these (much talked about) results, you will see that a lot of reporting tools are "self-enforced". Meaning, that you website administrators have to place a counter on each page to get an accurate account of what is going and what version of IE was used.

And now for the big problem for these results: It completely ignores the vast majority of the corporate world for the following reasons;

1. Large commercial organizations are simply not going to have "counter scripts" on their pages
2. Most web traffic coming OUT from an organization will have been filtered/blocked by a proxy server
3. Again, for large organizations, there is preferred (sometimes mandatory) browser - namely Internet Explorer. 

Taking these points together, I feel that Internet Explorer's browser share is significantly under represented by these types of analysis and stats. And, in particular my belief (from my daily experiences) is that more organizations than are reported in these types of metrics are using IE6.

Feeling courageous, I would hazard to add another 10% to the reported figures for Internet Explorer and  a similar increase for Internet Explorer 6.