Metric Madness
October 22, 2007
A story in today’s New York Times chronicles the difficulties faced by media companies in tracking online advertising. It seems that companies like Conde Nast, Viacom, and The New York Times itself are routinely finding that their estimates of how many people are viewing their advertising are higher than the estimates of independent groups like Nielson.
It’s an old problem, but one that the internet has aggravated. Because if you can’t show an advertiser just how many sets of eyes will land on their precious pitch, then you can’t easily extract the cash from their hand. And without advertiser cash, these media companies are dead in the water.
Add to the mix the metrics of ad delivery companies (ala doubleclick.net), and you’ve potentially got three different sets of numbers, all purporting to be an “accurate” measurement of ad traffic for a given site.
There are a host of reasons for the discrepancies in counts, ranging from viewers who delete sites’ cookies from their browsers to problems counting people who view the same site from home and the office. Still, a publisher’s estimation of its own traffic, when shown to greatly exceed the estimate of ad delivery sites and independent trackers, has to look fishy.
