With more and more of Net Applications's dirty secrets being dug up all the time, other stats companies are taking the opportunity to speak up and try to gather a following. One of them, StatCounter, has apparently been doing stats since 1999, but they don't seem to be quoted much by the media. …
While StatCounter's stats are probably as unreliable as everyone else's due to the endless possible error sources when doing browser stats, they have some interesting numbers that even seem to confirm just how heavily US-centric Net Applications is:
- They claim that Opera has 2.93% worldwide from January 1, 2009 to March 4, 2009
- Things aren't looking so good for Opera in North America, with a mere 0.47% in the same time period
- However, they claim that Opera has a market share of 7.27% in Europe
If we compare StatCounter's November 2008 stats to XiTi Monitor's November 2008 stats, they don't quite seem to agree on the numbers, showing Opera at 7.77% and 5.1% respectively. Could this have something to do with XiTi Monitor leaving out countries in Europe where Opera is doing well?
It should be noted that both StatCounter and XiTi Monitor show growth in Europe from September 2008 to November 2008. StatCounter shows Opera at 6.2% in September 2008, while XiTi Monitor shows Opera at 4.24%.
Another interesting observation is that StatCounter places Opera ahead of both Safari and Chrome worldwide, with Safari at 2.59%, and Chrome with a market share at 1.46%, less than half of Opera's claimed market share. This probably has to do with Opera's high market share in Europe, as in North America, both Safari and Chrome are ahead of Opera.
I have not looked into trends much, but it seems that there is a lot of fluctuation. If we look at the past 6 months, Opera had 0.59% in September 2008, 0.47% in October, 0.51% in November, down to 0.43% in December. In January 2009 Opera isn't even listed in the North American stats, while it makes a return in February with 0.48%.
Can we make any conclusions based on these numbers? Well, one could say that Opera is doing much better in Europe than in the US. But with the unreliable nature of browser stats, one should really take these things with a grain of salt. Sadly, even journalists are blindly quoting various stats companies, particularly Net Applications, without doing even basic research on the validity of the numbers that they publish.
Is there any harm in publishing flawed browser statistics? One might think not, but when people accept these numbers without questioning them, it can influence public opinion, which among other things means that browsers that don't seem to have a lot of usage will simply be ignored by many Web designers. There's often an attitude of "that browser doesn't show up at Net Applications, so no one should even care", which I believe prevents other people (browser users) from giving it a fair go.
And what if Opera or other browser vendors were to make decisions based on flawed statistics?
We must be careful when looking at browser stats, whether it's Net Applications, StatCounter, XiTi Monitor or someone else. If we blindly accept the numbers that are thrown at us (like many bloggers and journalists do), we are at risk of making the wrong decisions, and basing our attitude and behavior on flawed numbers.
In the end, that is not a good thing.