Yahoo Smacks comScore for Crap Data

Last Updated Aug 2, 2010 1:29 PM EDT

The high tech industry and those who watch it have become so enamored with facts and figures from comScore (SCOR), it seems to have become the official record keeper for everyone's Internet activity. There are just a couple of problems: all data is suspect, particularly if you can't find out exactly how someone created a forecast, and publicly-held comScore has a vested interest in being seen as the place to get the figures. So it was interesting today to see often embattled Yahoo (YHOO) smack down comScore for being enormously off in its estimations.

The action has implications for every manager in high tech who looks for hard data on which to base decisions and invests in reports from comScore. After all, if someone might be wildly off with one of the most popular sites on the Internet, how accurate will it be with every other company â€" and will you ever hear when it's not?

Yahoo has received many lumps during the tenure of CEO Carol Bartz, who has proven how combative she can be. However, there are times when being tough and throwing elbows is the only option, and this was one.

According to Yahoo, comScore miscalculated the company's traffic statistics last month. Not by a little -- a lot.

Yahoo! Inc. today announced that comScore has notified the company of an error in its June 2010 publication. comScore, as a result of this error, underreported Yahoo!'s U.S. page views by more than one billion and its duration metrics by more than 850 million minutes. Based on the corrected numbers, on a month-over-month basis, Yahoo!'s U.S. page views were down 4.7 percent versus the reported 7.4 percent, and U.S. duration was down 4.3 percent versus the reported 6.4 percent. comScore has committed to issuing the correct numbers in the company's next data release.
Woah! Those are some big errors -- and, apparently, comScore has admitted to a problem. Below is a chart that shows the before and after picture (click on the chart for the full size version):

But this isn't about Yahoo, which hardly comes out smelling like a rose, as U.S. page views and time spent on the company's sites per month are still down. However, forget that for a moment. (I'm sure Bartz would appreciate it.) Instead, think of the bigger lesson. Companies like comScore generally work with supposedly representative panels of consumers. But samples can go awry, especially if they don't represent the population as a whole. If comScore gets Yahoo wrong, then how reliable will the data be on less well-traveled sites? It's enough to make an executive rethink whether to slap traffic numbers on a business case projection.


Image: user COBRASoft, site standard license.
  • Erik Sherman On Twitter» On Facebook»

    Erik Sherman is a widely published writer and editor who also does select ghosting and corporate work. The views expressed in this column belong to Sherman and do not represent the views of CBS Interactive. Follow him on Twitter at @ErikSherman or on Facebook.