It's On! StatCounter Defends Browser Usage Share Data

In response to Microsoft's denunciation of a recent report in which Google Chrome allegedly took the top spot for one day over the past weekend, the company at the center of this controversy on Thursday issued a refutation of its own. StatCounter says that its data is valid and questions why Microsoft ignored some pertinent facts.

You may recall that StatCounter claimed that Google Chrome briefly overtook Microsoft's Internet Explorer to become the number one browser, if only for a day. That same day, Microsoft complained in a blog post that StatCounter's numbers were inaccurately gathered, in essence questioning the web metrics firm's credibility. (If not, I wrote about this event in Microsoft: Chrome Was Never Number One, Despite Reports.)

On Thursday, StatCounter defended its methodology.

"We disagree with Microsoft on a number of points," a StatCounter representative told me. "Chrome overtook IE in our stats for the first time last Sunday. The victory was narrow and short-lived but notable all the same for being the first time that this had happened."

According to StatCounter, Net Applications has itself only recently stopped measuring browser pre-rendering, a process that Chrome uses to pre-load web sites the user may want to visit next, speeding performance but, Microsoft claims, skewing usage statistics. "This point is not really relevant in comparing our stats at the moment," the StatCounter representative noted. "The really significant thing, however, is the undeniable trend in our stats towards Chrome usage at the weekend at the expense of IE."

StatCounter also countered a Microsoft claim about the geo-weighting of data, stating that it allows individuals to filter the overall usage numbers by country is desired. But Net Applications' data inflates IE usage in other ways, for example by including third party browsers that utilize the IE rendering engine. Stat Counter measures those browsers separately. 

"We try to be open and transparent about our stats," the representative added, questioning some of the other "fundamental differences" between her firm and Net Applications. For example, StatCounter uses a sample size of over 3 million web sites, while claiming that Net Applications gathers data from just 40,000 sites. Too, Net Applications only counts JavaScript-capable web browsers, while StatCounter measures both JavaScript enabled and disabled browsers.