A Must-Read...Part IV in a series of articles from Steve Rivers with Todd Wallace.

Steve Rivers: As we both know, when it comes to music research design, the screener has to be constructed properly to be effective. Otherwise, it?s garbage in, garbage out. What tips can you offer programmers in this area?

Todd Wallace: I think, in our respective travels as trouble shooters, we?ve both witnessed some music research samples that are more full of shit than a Christmas goose! And you?re right, snorkeling for the right respondents really is one of the most important considerations when it comes to music research.

The first thing you want to do is make sure you have a sufficient number of your P1 core partisans represented, since those are the listeners who are going to give you the most greatest TSL lift. Just as there are different horses for different courses, your competitive environment, and what you?re trying to achieve in your programming action plan for the near term, should dictate the ultimate composition of P1s in your sample.

In some cases, especially where you have the luxury of being the only station in your format, you may want to test only P1s. The downside to that is that it may inhibit cume growth, since you?re preaching only to the choir. But the upside is that you?ll end up super-serving your core which usually more than compensates for any mistakes you might make trying to expand your cume and, of course, there are many other ways to grow your cume.

In a direct head-on battle, I generally like to make sure that the sample is made up of a 2:1 ratio of core to cumers, which is usually the way the Arbitron P1 hierarchy stacks up. The important thing to remember when including other stations? core listeners in your sample is to make absolutely sure they indeed do cume your station on at least a weekly basis. You don?t want phantom cumers in your sample. Nor do you want non-cumers. That will only serve to contaminate your overall pool. But it?s amazing how many stations ignore this rule, they?ll just accept respondents willy-nilly, based on an Arbitron cume duplication table, with no forethought about whether they pass the OKOP test (Our Kind Of People).

SR: Reading the results of any research has to be done the right way, otherwise you could be chasing away listeners. I?ve always tried to look for songs that appeal to the core and the cume with a high degree of passion. Any thoughts?

You?re right on the mark. I?ve always believed, and preached anytime a soapbox was near, that favorite ness rules! And therefore, in my opinion, it has to drive all phases of your interpretive process. My way of dealing with that has been to add several proprietary evaluators to my MARS system computer program, so that they are there automatically for you, every time.

The first one is what we call the Skewed Average. Basically, it enables you to keep three eyes on your target audience at once. You always want to keep one eyeball on your P1 core listeners, the ones who listen to your station the most. At the same time, without taking your eye off your core, you also want to keep an eye on your total target including your cumers who also listen elsewhere. And finally, you want to keep an eye on your long span listeners, who contribute the most quarter-hours to your TSL. Then, we combine these three cross tabs, the total tab, your station?s core tab, and long-span TSL tab, into one enhanced cross tab that extends your overview with a positive skew.

The next evaluator is what I call the Hit Pattern Calculus, which is perhaps one of my biggest breakthroughs in music research. What I did here was to quantify all of the ingredients I found myself looking at in the interpretive process, rate them on a Bo Derek 10-scale in terms of what constitutes a true hit, and then factor them into one score that is more meaningful than any other reading. We do that for each song and then provide various rank-order summaries that make it more meaningful. This is the real bottom-line, so to speak, because it really simplifies the process since it takes so many factors into consideration. Using the earlier ?eyeball? analogy, it enables you to keep about 20 eyeballs on the audience. And it?s all driven by listeners ?passions?.

Then there?s an essence-testing cross tab I call Def-check, which is short for Definition-check. We target up to three defining songs that any or our true core listeners should like or love and offer a cross-tab of only those listeners who fit within that criterion. It?s just one more way to size things up, which can often be very illuminating. Frankly, I think some stations go a little bit overboard when it comes to defining compatibility and we?ve both seen instances where some research consultants have literally ruined successful stations by insisting on an overzealous adherence to this kind of evaluation. But, properly used, it?s another useful tool to help you define your station?s essence every week.

And finally, our MARS system?s Precision Dayparting Control lets you isolate cross tabs for up to dayparts to help shine a light on all those dayparting decisions or questions. It?s funny, what we?ve actually experienced over the years is that this PDC function usually results in less dayparting on a station, since it shoots holes in many of the old dayparting myths that were based on wives tales and hearsay. And that?s a good thing. Basically the message here is: in my experience, a hit is a hit is a hit, no matter what time of the day or night it gets played, so the less dayparting you do, the better.

SR: Right on the money. Play the hits. You are one smart mother Todd. Thank you for sharing so much of the true history behind callout. Hopefully, it?s taken some of the mystery out of the process.

Todd Wallace launched his consulting practice, Todd Wallace/Associates in 1975 and his track record of success-stories includes over 100 radio stations in the U.S., Canada, Australia, New Zealand, the Philippines, and the UK. In addition to being internationally recognized as the founding father of ?callout? music research, he is one of the pioneers of modern radio programming and format theory, having introduced such concepts as weekly tracking research, the ?P1? level of preferential listening, ?phantom? cume, and many more. His research company, Radio Index, has conducted perceptual and tracking research for some of the largest stations in America (including KIIS-FM, Kiss 108, WGCI, and many others). He also publishes a programming theory and promotional ideabank newsletter, Programmer?s Digest (with subscribers in 22 countries). Reach TW by e-mail at ToddWallaceTW@aol.com or by phone at 623-362-TODD (8633).