Chrome still trying to catch Firefox; IE still dropping, says Net Applications

At the end of 2011, it looked like Chrome had knocked Firefox from the number-two spot in Web browser rankings — but it’s starting to look like that title might have been a little premature. According to Net Applications’ Net MarketShare report, both Internet Explorer and Google Chrome lost ground in February 2012, leaving Firefox in the number-two spot with a two percent lead over Chrome — at least in the desktop browser market. All versions of Internet Explorer still account for a little more than half of all desktop browsers out there (52.84 percent, a drop of 0.12 percent since January), while Chrome slipped to an 18.9 percent share, down one percent from a peak of 19.11 percent in December 2011.

Apple’s Safari desktop browser also saw significant gains in February, jumping from 4.9 to 5.24 percent of the browser market. Opera also moved up from 1.67 to 1.71 percent.

The new figures highlight some of the discrepancies in how different firms account for browser share. Net Applications collects its data from customers using its live stats services, and says the data spans about 160 million visitors a month. Of course, there’s no way of knowing whether the sites using Net Applications services are representative of the Web as a whole; in fact, they’re probably not, since 76 percent report they participate in pay-per-click programs and 43 percent claim to be commerce sites.

Other services with different methodologies produce different numbers. For instance, StatCounter shows very different figures, with Internet Explorer accounting for only 35.75 percent of browser usage in February 2012, Chrome in second place with a 29.84 percent share, and Firefox in third with a 24.88 percent share. According to StatCounter, IE is in decline, Chrome is growing fast, and Firefox did see a slight uptick from January to February.

Why the discrepancy? Are one or both of these just doing it wrong? No — both companies have a solid history of reporting as accurately as possible using the data they gather. The issue is that Net Applications and StatCounter aren’t counting the same people. StatCounter is a Web analytics service, with customers using its code on more than 3 million Web sites around the world accounting for over 18 billion hits a month. If users visit a site with StatCounter code, the visit is included with aggregate data about browser share and other things. Again, there’s no way to know whether StatCounter’s base of sites is in any way representative of the broader Internet — and, in fact, they’re probably not since they’re sites looking for (and using) third party analytics services.

The bottom line: the fact major metrics services aren’t even in general agreement over their figures points to how difficult it is to measure browser share in the worldwide market. As always, consider such figures general guidelines, and consider any announcement of watershed changes in browser share with a grain of salt.