PPM Yays From The PPM Nays
May 14, 2012
After I participated in the PPM panel at the All Access Radio Summit, a fellow researcher forwarded a tweet from yet a third researcher that challenged my assertion that the PPM sample administered by Arbitron is not a representative sample. Gotta say, I hadn't thought that was going to be terribly controversial, since I figured most of us assumed that already.
For the last few weeks, NuVoodoo has been sharing empirical data showing exactly how those who say they'd consider carrying a PPM are different from the population at large. And remember, we're just looking at how the PPM-receptive universe compares to the generally research-receptive universe. Believe it or not, those who don't respond to research are different too! We had initially thought we'd do three columns spotlighting those differences, based on our 1,000 person national study. We've already done 5 and there are more to come. But the purpose of today's column is to discuss research in general and common sense in particular.
So, with that in mind, think about a few things you already know that will automatically differentiate the PPM yays from the PPM nays:
- People who don't care about radio won't participate (good news for radio)! This is a point of differentiation.
- People who don't care about the dollars that you get for schlepping that thing all over the place and having to move it every 20 minutes probably won't participate. This is a point of differentiation.
- People who don't want to be annoyed by having to deal with another gadget to carry around probably won't participate. This is a point of differentiation.
- People who are technology averse probably won't participate. This is a point of differentiation.
- Fact is, do YOU know anyone -- outside of the radio business of course â€“ who would carry that thing around? This would also be a point of differentiation.
The list goes on.
A number of years ago, when PPM was a mere glimmer in Arbitron's eye, I had the opportunity to do an experiment. Loading a predictive dialer with millions of PrizmNEâ„¢ coded and geographically-balanced phone numbers, we set about to do a study in 45 major markets over a period of months. After 2.5 million dialings and 30,000 completions, one might deduce that when you PrizmNEâ„¢ coded the completions, the percentages of each of the 67 life groups would be identical to the sample that went into the system in the first place. That would be a representative sample. Let's just start with the people who bothered to answer the phone. Right. They were different from the population at large. And then, when you filter down to people who said they'd consider participating in a PPM-type study, they're even more different. Now consider that people who say they'll do something are different from people who will actually do it. I think you get the idea.
Fact is, there's bias in any sample.Â Need empirical proof?Â Look at what happened to Smooth Jazz, Oldies and Classic Rock when markets moved from diary to PPM methodology.Â Do you think those formats suddenly just died?
Is this brain surgery? Of course not. It's common sense. It doesn't make sense that the PPM sample reflects the population at large. These are the cards we've been given, and the opportunity we have is to spend time finding out what makes PPM prospects and participants tick and aim our radio product and marketing squarely at them.Â That's the whole game.
I'm still not quite sure what to say to the research guy who actually thinks that PPM responders are just like the rest of us. But I'd suggest that radio folks think twice before engaging him to help win a PPM battle.