• by Joe Mandese  @mp_joemandese, March 12, 2008

What started out as a regularly scheduled meeting on Tuesday to update Nielsen clients on a variety of initiatives, erupted into a debate over the validity of Nielsen's plan to begin using its "live" national sample to test the simultaneous measurement of both TV and online viewing this fall. During the meeting, Nielsen said it has targeted a TV/PC sample of 375 households this fall, and would use that experience to evaluate a "broader deployment in the currency panel" by July 1, 2009.

By "currency panel," Nielsen executives were referring to the accredited, national TV sample it uses to generate TV ratings that are the currency for TV advertising buys, and are the basis for much of TV programming and scheduling decisions.

Nielsen previously announced plans to conduct the unusual and controversial test on its "live" sample, but Tuesday's meeting was the first time a group of high-level TV researchers reacted strongly to the plan.

"There are clients who want us to move faster and others who think we are moving too fast and don't want us to test in a live panel," acknowledged a Nielsen spokesperson. "We have to keep pace with the changing demand of the marketplace while protecting the quality of the sample and we think we've achieved that balance."

He added that Nielsen's plan is to conduct the test in careful phases and to monitor the impact each step of the way.

But some of researchers believe the magnitude of this live test is unprecedented and that the potential impact on the its national TV ratings sample is so profound that Nielsen ought to rethink the plan, especially given initial tests conducted outside its live sample that indicated a high degree of resistance among households to allow Nielsen to measure both their TV and PC usage.

"No research has been done by Nielsen to determine impact on ratings. It's called a 'live test,'" said one attendee. "The TV ratings currency will become a guinea pig or lab rat" that would be used to determine: the cooperation and fatigue of respondents and the impact on TV ratings.

The test panel would be a subset of 375 households within Nielsen's national sample of 14,000 households.

Nielsen is pushing forward with the plan, which is a critical component of its so-called A2/M2 (anytime/anywhere) initiative, and CEO Susan Whiting is said to be adamant that it must go forward.

But some researchers question whether there even is sufficient TV program available online to make the test comparable with conventional broadcast, cable and satellite TV program distribution.

According to the rules being used by Nielsen, such audience estimates will contribute to the "currency" - the ratings used to buy and sell TV ads - only if both the program and commercial content are identical in the online and offline versions of the shows. As Nielsen noted in its presentation, This is rarely, if ever the case with Internet delivered programs."