mozilla tech

Central Europe marketshare situation – short analyze (part 0)

My recent focus is on browsers situation in Europe. Especially non-western europe.  It means pretty big area with tons of countries, languages and home-grown Internet booms in the middle. I’d like to share some basic data about 5 of those basing on the data from Gemius.

Gemius is a polish company that focus on Internet statistics, profiling users for websites etc. What is good for us, is that the company is presenting weekly their merged results for “whole polish Internet” on the website Which gives everyone in Poland pretty good idea on what’s going on here.

Some time ago Gemius expanded to new markets – Czech Republic, Hungary, Lithuania, Ukraine. Another thing is that due to the rising problem of “cookie deletion” the ambitious methodology of tracking “visits” had to be shifted into “tracking page views”, and I want to present you the summary of this data in case of web browsers that I think is representative for those countries and this part of Europe (1).

In the next posts I’ll be analyzing each of the countries and in summary, I’ll present the combined data for the whole region.

1) To explain why I tend to assume that Gemius data for Poland is a good estimation for population, let me show same data.
According to InternetWorldStats, Poland has around 11,5 mln Interent users. According to last Gemius weekly stats before they changed the methodology (May 2007) they tracked 19 296 045 visits during the week.
Even if we assume it’s not 100%, it’s still pretty good sampling, especially as my goal is not to measure the amount of users, but the market share trends.
For Czech Rep., InternetWorldStats says they have 5.1 million users, and Gemius in May tracked 8 million visitors during the week.
For Hungary, IWS says about 3 mln users, and Gemius tracked 3.6 mln visits.
For Lithuania, IWS says about 1.2 mln users, and Gemius tracked 1.3 mln visits
For Ukraine, IWS says about 5.3 mln users, and Gemius tracked 1.2 mln visits.

Usually, one user is doing far less than 3 visits per week which means that you can divide the visits by 3 and get estimated real users. Comparing IWS Internet users to those real users will give you a sampling.
In all those cases, beside Ukraine, we can say that it’s very representative sampling. In case of Ukraine, I still tend to believe that it’s enough to get some idea on what’s going on there.
please, bear in mind, that those are statistics. Statistics are a way to represent the data, and data represents reality. On each level there are simplifications and mistakes, and it’s up to you to decide on your own how far you trust the summary results.

mozilla po polsku tech

Tak się robi historię…

Przepis jest prosty, bierzesz historię i ją robisz. Korzystając z faktu, że Tobie zależy na pewnej jej wersji bardziej niż innym, nie napotykasz raczej oporu, skutek jest wszystkim.

Ostatnio często mówi się opoprawianiu Wikipedii“. Jak wiadomo, Wikipedia nie jest doskonała, i można ją poprawić. Nasz wewnętrzny dysonans poznawczy wręcz marzy o takiej okazji zewnętrznej racjonalizacji naszych przekonań, decyzji i wyborów.

Niestety nawet na naszym, małym, podwórku przeglądarek internetowych nie każdemu udało się oprzeć tej pokusie “wykreowania” kawałka historii.