Monday, July 30, 2007

Arbitron Advisory Council Reads And Weeps

(click on the images to enlarge them)

Arbitron, how can this possibly BE? I thought, based on what we had been told by your top level executives that in exchange for that 60% rate increase you'd be tracking compliance on a daily basis and replacing panelists immediately if their use of the PPM started to fall.

PPM markets, get ready for sample buffering just like we currently see in diary samples. It looks to me like they need to start replacing fatigued respondents in week 4, week 8 and weeks 11 and 12 just like they often do in diary markets now.

Given that possibility, you have to ask was this really worth betting the company on? I know it's the first inning of the game and we're all just learning about the process. But, hasn't Arbitron been testing and sampling PPM for a decade or more?

Looking at these charts, I wonder if it would have been wiser to just invest the rate hikes we're paying for PPM in larger diary samples. Weighting, creating unacceptable and unbelievable wobbles in many western markets, for example, this spring has been at unprecedented levels. PPM was supposed, we had been led to believe, to improve compliance and sample stability, reducing the need for sample weighting.

After all these years of development, this info - now that three markets will be dependent on the new technology by the end of 2007 - is more than a little surprising and disappointing to see.


Frank Saxe said...

Arbitron agrees "there's absolutely churn" - CEO Steve Morris says PPM panelists tend to drop out in the first few months of being on the panel. He says "Our experience of compliance in Houston is a pretty good indicator of how that will go." He tells a recent conference call that they've begun increasing incentives to certain demos to see if that will make a difference.

Dave Lange said...

Hopefully Arbitron will not just invest more into the diary. Going back to a memory game as we sell against the new 'metrics' of the web where they can tell you who, what, when and where and maybe even why for every one who saw your ad with NO recall error. We've already seen less respondent bias and giving it up because of some sample churn is minor. The big concern we all should have is the 'have PPM' and 'have NOT PPM' markets. How will places like Madison Wi be taken seriously by the clients there as they wander around with old/tired data collection in a soon to be 45 million dollar radio market?