Web Browsers: the War of Numbers

Web browser usage statistics have existed for years and each month, it is possible to follow their evolution. For this, several services do their analysis and publish monthly reports. These sites also adapt to the evolution of the Internet as a whole with, in particular. Statistics on mobile browsing or even on social media market shares. But each month, an underground battle begins between the various analysis services. Sometimes releasing completely different figures.

The two major sites are Net Market Share and Stat Counter Global Stats . Which have diametrically opposed methodologies and therefore significantly different results. Simply by taking a look at the methodologies adopted. Which makes approximately 1300 unique visits per day.

Web Browser Usage Statistics

 

The service also weights the figures by country. Stat Counter is based on a sample of 3 million sites by counting the number of page views. Three criteria emerge from these methodologies: The sample: It is important to have the widest possible Singapore WhatsApp Number sample to get as close as possible to reality. Net Market Share only has 40,000 (for about 160 million unique visits) against 3 million for Stat Counter. Unique visitors vs. seen pages : Big cleavage between the two sites on this criterion.

For Net Market Share, the number of unique visitors is much more representative because it is less subject to results biased by page views and possible bots. Stat Counter defends itself by criticizing the fragility of a measure on unique visitors: on what basis? Cookies ? The IP address? What about dynamic IPs? According to them, the pages seen would be a much more reliable indicator of a tendency to use. one browser rather than another (especially in the case where several browsers are used).

Stat Counter Global Stats

Singapore WhatsApp Number
Singapore WhatsApp Number

 

Weighting by country: Here again, a major point of contention. Net Market Share performs it on its figures and in particular for China which, by its imposing population, weighs a lot in the market shares. Stat Counter decided not to carry out weighting and therefore to put all the countries at the same level, without proportion according to the populations. The site justifies itself by criticizing the unreliability of the various weighting systems available and leaving the user the possibility of doing it himself.

The two sites also disagree on using CIA data for this weighting, with Stat Counter arguing that it’s outdated and doesn’t distinguish between mobile and desktop users. The discrepancies are marked, as are the differences in results. Both camps have their arguments and each believes that its method is better than the others. A Wikipedia page lists the figures of the various statistical study firms and their methodologies.

Leave a comment

Your email address will not be published.