Web measurement players put to test

By Published on .

Web measurement companies never agree how many people visit a site or click on an ad, but they have agreed to explain their differences.

Top players will take part in an experiment to measure and monitor one online campaign, with results released Nov. 1 at the @d:tech conference in New York.

Measurement "is probably the most important issue facing online media today, and . . . is probably the most misunderstood," said Rich LeFurgy, chairman of the Internet Advertising Bureau and a member of the @d:tech advisory board. "What this panel will do is illustrate that we do have very good measurement in place, contrary to popular belief."


Three Web measurement suppliers--Media Metrix, Nielsen/NetRatings and AdRelevance and--will track an Advertising Council banner campaign from Saatchi &|Saatchi to promote the Points of Light Foundation. DoubleClick agreed to deliver 2 million impressions via its network. Millward Brown Interactive will study brand impact. Engage Technologies' I/PRO will review all the results.

The panel will include side-by-side comparisons, such as Millward Brown and Nielsen/NetRatings home-user demographics. But @d:tech Chairman Kate Maddox said the bigger goal is to show what can be measured and strengths and limitations of various approaches.

"I think it's a question of interpretation," said Ms. Maddox, a former Advertising Age Interactive editor who is a VP of @d:tech's producer, eMarketWorld.

I/PRO VP-Marketing Tim Reed said each service will be asked to comment on a specific area. "We're presenting it more as a flow than a battle of the titans."

Wouldn't life be easier if there were one, accepted data source?

"People will understand there are differences," said Mr. LeFurgy, a venture capitalist and investor in start-up AdRelevance. "It's not meant to say that one number's right and another number's wrong."

That leaves it to the advertiser to decide who measures up.

Copyright September 1999, Crain Communications Inc.

Most Popular
In this article: