I also figured that their data is biased, but the interesting part is that the trend line goes counter to that bias. Say that RescueTime is heavily used by early adopters that are comfortable with cloud-based services, the same people that are likely to use Google Apps. In the beginning, you'd expect that to bias the results toward Google. But as RescueTime broadens toward mainstream appeal - which is already happening, I was an early adopter but then dropped it when it got too corporate - you'd expect Google's market share to drop, as RescueTime's share expanded to include more Microsoft users. Instead, you see the opposite, where Google's share is growing even though the population of RescueTime users is expanding into domains typically held by Microsoft.