OpenOffice.org by the Numbers
What would you say if you heard that OpenOffice.org had an impressive 22% share of the market? It does, in Poland and the Czech Republic. At least according to numbers pulled together by Webmaster.de using FlashCounter. The analysis looked at visitors from more than 20 countries and found use of OpenOffice.org is highest in Europe and not even in double digits in the United States.
The results may not be entirely accurate, but at least they provide a picture of office suite usage. The method used was to examine fonts installed on systems to identify various office suites. The OpenOffice.org numbers roll up all variants of OpenOffice.org — IBM Lotus Symphony, StarOffice, NeoOffice, etc.
OpenOffice.org enjoys a share above 20% in Germany, Poland, and the Czech Republic, and above 15% in Spain, Italy, and France. It doesn't quite crack double digits in the United States, though, with only 9%. Canada favours OO.org a bit more, with 11% share. OpenOffice.org is only a blip in China if the stats are to be trusted, giving only 0.2% of the market to OO.org in China while MS Office weighs in with 68% there. The study also tracks iWork, and finds it has no adoption above 4% in any country.
The study also cites a strong correlation between OO.org share and Firefox market share. Not surprising at all, and the two projects should probably think strongly about doing joint marketing on an ongoing basis.
According to the report, the error margin could be as much as +/-5%, which is a pretty large fudge factor. The study also fails to take into account actual usage and users who may have multiple suites installed. Many people using Linux, for example, have OpenOffice.org installed but may be using something else entirely, like Google Docs, or may not be using an office suite at all.
Market share analysis of open source software is fairly tricky. The OpenOffice.org project collects some data based on analyst research, download stats, and so on. Download statistics are wildly inaccurate, though, because the numbers include individuals re-downloading OpenOffice.org, different releases, and the fact that the downloads don't reflect copies distributed through other means.
Most Linux users get OpenOffice.org via their distribution, and OpenOffice.org has several offshoot projects that could be considered for market share purposes but not trackable by OpenOffice.org itself. In short, nobody really knows how many users of OpenOffice.org there are, but the FlashCounter tracking at least attempts to measure installed versions of OpenOffice.org on individual machines.
Taking the entire thing with an enormous grain of salt, it's at least good to see someone trying to nail down numbers. Given what I've seen of free software use, they seem to reflect reality. Given that OpenOffice.org enjoys almost no advertising budget, the results so far are fairly impressive. It should be interesting to see what, if anything, Oracle does differently in supporting the project and if that does anything to budge the numbers.