How did they do this study? "The survey compared total home and work usage minutes between March 1999 and March 2001. Total usage minutes includes the time spent at a particular
online property, Web site, category, channel or application." That's not very clear. Did they just get a sample of people to fill in a form estimating time spent? That would clearly skew the results in the "big 4" direction for two reasons:
1) It is easy to remember that you spent a lot of time at, say, Napster, but harder to remember all the other little sites. So if you are retroactively filling in a survey, I would think that the main sites would be over-remembered compared to the aggragate of all the little sites.
and
2) No sophisticated web users would take such a survey.
Or, if they are actually tracking people, then they are not really counting exact time spent, they could only be tracking clicks, and estimating time spent (I might click to MSN, and then go get a cup of coffee. When I come back 15 minutes later my computer is still on MSN, but I haven't really been "on" the site.) And worse than that (if they are doing it by click) is that microsoft redirects you to hell and back every time you hit one of their sites. Go to hotmail, for instance, and you'll be redirected to passport and then back to hotmail. Same thing with all their sites. They do this to try and consolidate all traffic through passport, but this means 2 or 3 "clicks" for every one click you make. Do they count all these. Or how about every Netscape browser that starts up by automatically contacting netcenter.com (AOL property.) Do these count, even though it's just the case that a lot of people don't know how to change the default start up page in their browsers?
I'm pretty suspicious of this study. Good link though.
as the article said (cnn is owned by time warner aol) so it could be biased
yes, i would be interested in the sampling methods as well. maybe they tracked via a cookie for a period of time on each person's computer. but i think you are right in that the people who would let someone track their internet acitivites might not be representative of all internet users. also where did they advertise to get the sample base - msn, aol? troubling.
i visited the jupiter media metrix site, hoping to get more details, but did not find much except the press release, although i think you may be able to request add'l info.
|
online property, Web site, category, channel or application." That's not very clear. Did they just get a sample of people to fill in a form estimating time spent? That would clearly skew the results in the "big 4" direction for two reasons:
1) It is easy to remember that you spent a lot of time at, say, Napster, but harder to remember all the other little sites. So if you are retroactively filling in a survey, I would think that the main sites would be over-remembered compared to the aggragate of all the little sites.
and
2) No sophisticated web users would take such a survey.
Or, if they are actually tracking people, then they are not really counting exact time spent, they could only be tracking clicks, and estimating time spent (I might click to MSN, and then go get a cup of coffee. When I come back 15 minutes later my computer is still on MSN, but I haven't really been "on" the site.) And worse than that (if they are doing it by click) is that microsoft redirects you to hell and back every time you hit one of their sites. Go to hotmail, for instance, and you'll be redirected to passport and then back to hotmail. Same thing with all their sites. They do this to try and consolidate all traffic through passport, but this means 2 or 3 "clicks" for every one click you make. Do they count all these. Or how about every Netscape browser that starts up by automatically contacting netcenter.com (AOL property.) Do these count, even though it's just the case that a lot of people don't know how to change the default start up page in their browsers?
I'm pretty suspicious of this study. Good link though.
- jim 6-08-2001 3:24 pm
as the article said (cnn is owned by time warner aol) so it could be biased
- Skinny 6-08-2001 6:41 pm [add a comment]
yes, i would be interested in the sampling methods as well. maybe they tracked via a cookie for a period of time on each person's computer. but i think you are right in that the people who would let someone track their internet acitivites might not be representative of all internet users. also where did they advertise to get the sample base - msn, aol? troubling.
i visited the jupiter media metrix site, hoping to get more details, but did not find much except the press release, although i think you may be able to request add'l info.
- linda 6-08-2001 7:36 pm [add a comment]
Methodology?
- alex 6-08-2001 8:00 pm [add a comment]