(or, an alternate title: ‘inferring trends from a single data point’ :)
Last night I had a look at the W3Schools.com browser data and saw that they had put up the July numbers… and I was SHOCKED to see that the IE7 numbers were DOWN:
Assuming for a moment that this is not an error, what is the cause? After the June numbers were posted, I commented “With the release of the final FF3 in mid-June it will be interesting to watch to see if this causes a significant bump in the July numbers.” Apparently it did; from June to July:
A point that should be noted: the IE7 to FF3 switch rate is bigger than that six-tenths of a percent indicates; IE5 computers are pre-XP and are increasingly being taken out of service (as well as older XP ones) and replaced with new ones that run Vista (and so only IE7) and machines running XP are slowly switching over from IE6 to IE7 via Windows Update. The outmigration from IE7 to FF3 was thus likely to have been just over 1%; that certainly is impressive. FF3 seems to really be a crowd pleaser (note the increased FF usage since March; that roughly corresponds to the late beta releases for FF3). Microsoft had better be sure that IE8 is damn good and can compete with not just FF3, but also FF3.1 which is due around the same time, if it wants to try to slow its market share slide.