20 October 2011

Predicting browser market share - 2011 edition

About a year ago I wrote an article about browser market share and the expected evolution for the coming months, using my own method based on Markov chains. Looking back now at the actual evolution, it’s pretty easy to see the prediction matches the actual trend only for a short while, a couple of weeks at best, then starts to fall behind, to underestimate the growth of and the shrinking market share of Internet Explorer. The forecast is relatively good for on the other hand, even after 6 months, meaning that the launch of version 4 in March didn’t affect the established trend of stagnation, followed by slow decrease. The model itself does tend to underestimate future evolution: it evens out growth and decay until all products on the market achieve an equilibrium. This means it’s not very well suited for long-term forecasts, because on longer time-frames the shifts in browsing behavior will accumulate and significantly deviate from the forecast.Browser market share 2010 forecast vs. actual evolution

Let’s see what the model will predict for the next couple of weeks and the beginning of 2012. I won’t go into the details of the method used anymore, since they are pretty well explained in the previous article and haven’t changed much. Naturally, this time around I used the worldwide browser market share data from StatCounter for approximately the last year and extended the forecast about 20 weeks into the future. As other articles pointed out, the market is nearing the point where Chrome will become the second most-used browser, overtaking Firefox, who held that spot for years. According to this forecast, this should happen around week 48, at the start of December. About at the same time, in week 47 (give or take 2-3 weeks for random variations), Internet Explorer’s share should slip below 40%. At the end of 2011, the predicted market share distribution would be: 39.3% for Internet Explorer, 26.1% Firefox, 27.3% Chrome and 5.1% Safari. After that, in another 10-12 weeks, Chrome should already pass the 30% mark, if the current trends don’t change too much. Browser market share forecast end 2011

As mentioned last time, the Markov model offers some insight into the patterns of user migration from browser to browser, for example:

  • The number of users who stick with Internet Explorer on a month-to-month basis has decreased, now estimated at around 97.2%. The rest will migrate to Firefox. The model also shows around 5.8% of users migrate to Internet Explorer from Firefox or ‘other browsers’ ( & co.), a percentage that has decreased since last year, explaining it’s accelerating decline;

  • Only 95.3% of Firefox users will be using the same browser the following week, while it attracts a small percentage (4.4%) from ‘other browsers’ and Internet Explorer. Unfortunately for Firefox, it has the lowest loyalty among the top four browsers, so getting back to the second spot and to a positive trend looks to be very difficult;

  • Chrome has by far the most loyal users: none of them switch away after adopting Chrome, the same result as last year! So the current growth is expected to last, fueled also by ‘converts’ from competing browsers (about 4.7% each month).

It’s interesting to notice how the current top browsers are both in decline. It’s not that easy to find some common reason for this situation, a trait that both share, a weakness that Chrome exploits, because Firefox is a modern browser and has been long before Microsoft remembered to compete and try to offer a decent product. The only thing that comes to mind is the internal fragmentation and the slowness to update users to the newest version; a relatively new problem for Firefox, but a long-known issue for Internet Explorer. Could these two browsers cannibalize themselves so badly, that Chrome becomes the top browser by the end of next year?

Post a Comment