Localization Statistics

This page is the home of an effort to determine one possible metric of browser localization coverage - that is, the percentage of the Internet population who have the browser in their first language (a.k.a. "language spoken at home"). This is not the only way of measuring, but I think it is one valid one. The higher the percentage is, the more people you are giving the best service to.

Calculating this figure is difficult because available data for "Internet population" is given by country, and not by language. So, I have taken the net population figures for each country and split them up proportionally to the first languages spoken in that country. There is an assumption here that the internet population of a country is evenly distributed among its language communities, which is probably not true for a country like India. However, this method will give better results than just allocating languages to countries on a 1:1 basis - "USA - English; India - Hindi; Italy - Italian" and so on. (Did you know that only 82.1% of the population of the USA speaks English at home?)

Better data is always welcome - and if any country can provide data of internet population split up by language, that can be incorporated instead. But as yet, I have found none.

Download the spreadsheet (OpenDocument format).

Headline Coverage Figures of World Net Population

Total World Net Population: 1,582,257,085

Firefox 3.05992.4%
Firefox 3.56895.7%
Opera 9.63584.8%
IE 77798.3%
IE 85794.6%


Why does the "Firefox 3.5" line say there are 68 localizations and not 75, when everyone's been blogging that there are 75?
This work is an overview, which comes with some simplifications. For example, I only have one column in my spreadsheet for each of English, Portuguese and Spanish (and some other languages like Tamil). That's not to rubbish the work of localizers like the en-GB team; but, as a Brit myself, I'm happy to admit that I could use en-US at a pinch, and it would still be much better than having to use a browser in Urdu or Simplified Chinese.
Which language communities do we serve which no other browser vendor serves (in their latest stable release)?
Hello to native speakers of Esperanto, Welsh, Persian, Irish, Galician, Icelandic, Kurdish, Occitan, Romansh and Sinhala. We're meeting your needs :-)
What do you define as a localization?
The above figures now only include officially-released localizations.
Are IE's localizations or Opera's localizations better or worse than the Firefox ones?
That's a question of some debate. My view is that not all localizations are created equal, but any localization is much, much better than no localization. In this spreadsheet, I lump together 90%, 95% and 100% solutions and call them all "supported", which is by contrast to the 0% solution, no localization, which is "not supported". I think this is a reasonable simplification, and it also avoids having to have arguments about whether or not an IE LIP is superior or inferior to a Firefox addon language pack.
If Microsoft were reading, which language should they do next for IE 8?
Tagalog, Persian, Kurdish, Urdu and Azerbaijani, in that order.

Original URL: http://www.gerv.net/hacking/l10n-stats/