ComScore and the ‘Traffic Assignment’ Problem

A week ago I wrote a post that showed IYP sites/networks were still doing well notwithstanding all the pressure they were under from Google. In making that assessment I relied on comScore’s Top 50 traffic tables from a year ago and December, 2010.

I said in the post: “Yellowbook wasn’t on the chart in 2009, now it leads the category — remarkably. Superpages has  improved its overall position and AT&T is in a very similar  position as last year.” My intention wasn’t to compare these sites to one another but to show that collectively IYP sites seemed to be doing OK.

I got a call and was “schooled” in the potentially significant inaccuracy of this traffic data. Publishers have forever disputed comScore’s accuracy. But I guess I’m naive; I never knew that traffic data could be manipulated through agreements between sites.

Apparently Site A can assign traffic to Site B, either because it was paid to do so or for no reason at all. Site B then gets credit for Site A’s traffic. ComScore apparently doesn’t independently penetrate or investigate these deals/assignments. It merely reports them as it’s told to do.

Accordingly some of data on the Top 50 chart may be highly inaccurate. I’ll report back from comScore when/if I hear something on this.

You can follow any responses to this entry through the comments feed.

3 Responses to “ComScore and the ‘Traffic Assignment’ Problem”

  1. Pankaj Mathur says at

    So what is the implication here in context of IYPs? Are they buying traffic or getting redirection from other sites?

  2. Chris Silver Smith says at

    Yes, it’s a fairly open secret that internet yellow pages companies have long bought some significant portions of traffic from other sites.

    Generally, when a YP company gets an agreement to distribute their ads and listings onto another site, they negotiate with that site to include the rights to get the traffic assigned to them in comScore and other reports.
    On one hand, this sounds like a trick of some sort to make themselves look better than they really are in terms of overall traffic. In actuality, I think this is more an artifact of the print industry where they originated, which always sought to use 3rd-party validation of circulation figures, similar to magazines, in justifying advertising rates. So, the assignment arrangement makes sense when they’re demonstrating to agencies and other parties how much distribution they have available.

    It sounds like the person “schooling” you felt somehow sensitive about the relative audience share being reported. If it was a yellow pages company, it’s sort of funny because they’re all on equal footing — they nearly all have the ability to buy traffic and negotiate assignment rights.

    It’s really comScore’s somewhat odd methodology that’s to blame for the situation. However, it really is an artifact of how ad rates are bought/sold/justified that also results in some of these oddities.

  3. Greg Sterling says at

    Yes . . . these numbers appear to be a distortion but one that cuts broadly across sites.

Leave a Reply