Quantcast

ComScore: Cookie Crunch Causes Crummy Counts

Get the WebProNews Newsletter:
[ Business]

Counting the online audience through the use of cookies may be overstating the size of the audience for websites, due to Internet users deleting cookies from their machines.

ComScore: Cookie Crunch Causes Crummy Counts
ComScore: Cookie Crunch Causes Crummy Counts
ComScore: Cookie Crunch Causes Crummy Counts

The usefulness of cookies has been debated virtually since the time they came into broader use online. Cookies are small text files stored on a person’s computer when they visit websites that use them. The website can in turn use those cookies later to enable certain commerce functions, or to track their movements for analytic purposes.

ComScore said in a study of 400,000 home PCs in December 2006 that cookie-based measurements could be off by 2.5 times, or 150 percent. That’s a serious overstatement with implications for advertising and audience measurement.

The regular deletion of cookies by some Internet users causes this inflation. ComScore’s Magid Abraham said: “With just 7 percent of computers accounting for 35 percent of all (first party) cookies, it

ComScore: Cookie Crunch Causes Crummy Counts
Top Rated White Papers and Resources
  • Adrienne Boswell

    I wrote something that looks at the server variables when a user comes on to the page. I look at IP address and UA string. IP addresses will most of the time be unique, except in the case of AOLers and others of their kin, and even with those, one can still differentiate based on IP+UA string.

    Cookies are bad not only because people clear them, but some people have their browsers not accept them at all, based on their privacy settings.

    It’s actually a crap shoot, but I almost never depend on cookies.