3 ms·

I found it interesting that the sum actually isn't relatively constant. The sum is steadily declining from 1.89M in februari 2009 to 1.65M in februari 2010. Giv

I found it interesting that the sum actually isn't relatively constant. The sum is steadily declining from 1.89M in februari 2009 to 1.65M in februari 2010. Given the ongoing growth of the internet (and the population), the total traffic should have increased, making the shrinkage all the more interesting. I'm guessing there are either Asian sites eating away a substantial market share or that the measurement methodology of Compete has been changing.

alanstorm 17y ago
My guess would be on their measurement methodology. The about page reads: "We have a diverse sample of 2,000,000+ U.S. Internet users that have given us permission to analyze the web pages they visit and ask them questions via surveys." That translates as sketchy toolbar/shareware installs and checkboxes that are checked by default.
- iron_ball 17y ago
  And that means this stat likely does not reflect the usage habits of technically proficient users, because they are the least likely to have allowed those grayware installs.
  - aaronbrethorst 17y ago
    Which would imply SO's traffic is dramatically under-reported. In fact, August they had a blog post that indicates Compete's data is off by a factor of 15, and this was 7 months ago! http://blog.stackoverflow.com/2009/08/one-year-of-stack-overflow/ http://blog.stackoverflow.com/2009/08/one-year-of-stack-over...
huhtenberg 17y ago
It just means that Google is getting better at serving correct A's for common Q's, and so there is less need for visiting specialized sites as these two.
mnemonicsloth 17y ago
Most of the drop occurs during the winter holidays, when people have less time to work. The biggest dip is reported for January and probably corresponds to numbers taken in December. That still leaves 12% decline between this February and last, though.
billybob 17y ago
Another possibility: SO just sucks less. It takes fewer page hits to find the answer you need. This is my experience, anyway.

alanstorm 17y ago

My guess would be on their measurement methodology. The about page reads: "We have a diverse sample of 2,000,000+ U.S. Internet users that have given us permission to analyze the web pages they visit and ask them questions via surveys." That translates as sketchy toolbar/shareware installs and checkboxes that are checked by default.

iron_ball 17y ago
And that means this stat likely does not reflect the usage habits of technically proficient users, because they are the least likely to have allowed those grayware installs.
- aaronbrethorst 17y ago
  Which would imply SO's traffic is dramatically under-reported. In fact, August they had a blog post that indicates Compete's data is off by a factor of 15, and this was 7 months ago! http://blog.stackoverflow.com/2009/08/one-year-of-stack-overflow/ http://blog.stackoverflow.com/2009/08/one-year-of-stack-over...

huhtenberg 17y ago

It just means that Google is getting better at serving correct A's for common Q's, and so there is less need for visiting specialized sites as these two.

mnemonicsloth 17y ago

Most of the drop occurs during the winter holidays, when people have less time to work. The biggest dip is reported for January and probably corresponds to numbers taken in December. That still leaves 12% decline between this February and last, though.

billybob 17y ago

Another possibility: SO just sucks less. It takes fewer page hits to find the answer you need. This is my experience, anyway.