Nielsen//NetRatings Spills Metric Guts

Nielsen//NetRatings (NNR) trades in information, metrics for Web
advertisers and publishers. And now the firm will tell us how it finds that information.

The Interactive Advertising
Bureau (IAB), which represents some of those who buy that
information, recently demanded that Nielsen explain its methods.

Today, the firm said it will open up its Internet audience measurement methods to a Media Rating Council’s (MRC) accreditation process. NNR also plans to continue an already-existing MRC research plan for
its panel procedures. The MRC is under the oversight of the U.S. Congress.

“The MRC process is the only audit that certifies to clients and to
the industry that we have fully disclosed our methodology and that we
are executing against that methodology,” NetRatings vice president
Manish Bhatia said in a statement.

NNR committed to an MRC audit of two of its metrics tools: NetRatings’ desktop meter, which collects in real time Web site visits, time spent and pages viewed by Internet users; and NetRatings’ patented page-tagging technology. NNR also plans to continue an already-existing MRC research plan for its panel procedures.

In late April, IAB President and CEO
Randall Rothenberg wrote an open letter to comScore President Magid
Abraham and Nielsen//NetRatings President and CEO William Pulver in
which Rothenberg said he wanted their companies to commit their
interactive-audience measurement processes to a third-party audit.

The problem, Rothenberg wrote, is that NNR and comScore numbers are
still based on small samples of Internet users.

The sample sets are called “panels” in the advertising industry, and
Rothenberg said their quaintness, recalling how in the 20th
century there were “small panels of Americans who allowed their TV
sets to be wired, filled in diaries, or sent prepaid postcards
listing their preferences back to the research firms in the Princeton-New York corridor.”

Quoting himself from an article in an 1998 issue of Wired, Rothenberg
explained how these panels fail to take advantage of what he has
always seen as the Internet’s promise: “its ability to tell, to a
high degree of certitude, how many people are coming to, perusing,
and engaging with a media outlet.”

Panel measurements, he argued, are the cause for discrepancies
between server logs and the numbers comScore and NNR publish. Those
discrepancies make it too hard for advertisers to determine if they
are getting what they are paying for at a time when they are
increasingly relying on the Internet to market.