It is easy to detect from the user agent what type of device the incoming request comes from.
Since nobody mentioned any specific measurement, I will use as an example the ones from StatCounter. They analyze user agent data from page hits in a couple million websites. If you go to their GlobalStats subdomain you see browser market share in a time series graph. Overall or platform-specific, that is already accounted. Even tablets are separate from phones despite running the same OS.
Firefox on the desktop has a global market share of 10% according to GlobalStats, while on the average between all devices it’s only 4%.
That data is flawed based on the fact that not data from the entire web is being collected (though 3 million common pages is a large enough sample size) and that absolute page hits are accounted rather than unique visitors. But, there is no reason that users of one browser would hit pages more often than users of the others. Anyways, those are the 2 facts that come to my head right now that may distort the true data.
And, as you can expect, you can’t reliably measure ‘market share’ of things that are not in the ‘market’.
It is easy to detect from the user agent what type of device the incoming request comes from.
Since nobody mentioned any specific measurement, I will use as an example the ones from StatCounter. They analyze user agent data from page hits in a couple million websites. If you go to their GlobalStats subdomain you see browser market share in a time series graph. Overall or platform-specific, that is already accounted. Even tablets are separate from phones despite running the same OS.
Firefox on the desktop has a global market share of 10% according to GlobalStats, while on the average between all devices it’s only 4%.
That data is flawed based on the fact that not data from the entire web is being collected (though 3 million common pages is a large enough sample size) and that absolute page hits are accounted rather than unique visitors. But, there is no reason that users of one browser would hit pages more often than users of the others. Anyways, those are the 2 facts that come to my head right now that may distort the true data.
And, as you can expect, you can’t reliably measure ‘market share’ of things that are not in the ‘market’.