A few weeks ago I promised in my MusicBiz column that I would seek out Todd Wallace. Todd?s the Godfather of radio music research. I wanted to get some insight into how he first developed callout. Todd and I have known each other for well over 25 years. We worked together as programmer and consultant on different projects many times, including my stint as PD at both KIIS Los Angeles and WXKS-FM Boston. He?s helped me fine tune my music systems and has always been a great sounding board for good ideas as well as creating effective ratings promotions. What follows is two-way dialogue about my interest in finding out how it all started.

Steve Rivers: Todd, you?re acknowledged to be the originator of callout music research. In fact, any callout system in play today is an offshoot of your idea. Tell me, when did you come up with the concept and where was the first place it was used?

Todd Wallace: It?s actually a long story that evolved over several years. The original germ was an idea I had been thinking about and developing when I was working as the young all night jock at CKLW in Windsor/Detroit. I wrote a memo to PD Paul Drew outlining how flawed request-line tabulations and record store sales were as an indicator of actual popularity (since there was a limit on how many songs a consumer could or would buy, and the fact that once they bought it, their .one vote. was already spent. In addition, neither requests nor sales took into account listeners ?dislikes?, or what we now call ?burn?.) I suggested to Paul that we might want to do some in person record store research to ask record buyers what other songs they would also buy if money was no object. Also, going further we?d have our request line operators ask about other favorite songs and if there were any songs which callers did NOT like. My reasoning was that most listeners, especially adults, were a ?passive? majority that often didn?t get ?counted? or heard from because they didn?t regularly purchase singles or didn?t have the time to call radio stations to make requests. Well, that was pretty revolutionary stuff for 1969, so Paul politely told me it was an idea way ahead of its time and we left it at that. But he encouraged me to keep developing the concept on paper, which I did, hoping for a chance to implement it somewhere. Anyway, that?s where the idea was born.

Flash forward to KRUX in Phoenix 1971, where I had become morning DJ and PD. GM George Lasley was intrigued by some of my ideas on how to get to know our audience better. Back in those days, there was no budget line item allocated for ?research? since most radio stations didn?t do it. But, I was able to get a few dollars of slush from other budgets and we figured out ways to do the proverbial ?more with less? on several levels.

We commissioned a local university?s marketing class to do a random survey of key perceptual questions as a mall intercept project. The results were among the first of their kind and surprisingly accurate (even by today?s standards). It was very revealing insight into how listeners actually listened. Some of the most seminal truths of ?radio listening basics? were uncovered in that study.

We also conducted several what we called, ?Listener Advisory Board? luncheons, which were like informal focus groups. We discussed every aspect of what we did on-the-air with average listeners (especially teens), asking what they liked and didn?t (and why) and taking note how they described things. Those too, were eye-opening, because it added another camera angle to any fuzzy statistics and most of all it clearly confirmed the ?passivity? of the audience that I?d long suspected.

We also pioneered weekly tracking research at KRUX. One of the DJs on the staff was the perennial ?Chicken Little?, always worrying that the sky (and our ratings) were falling if he didn?t hear the station blaring everywhere he went. Well, after experiencing one too many of his paranoid rants, I figured that doing some quick and dirty random callouts would give us ?instant ratings? every week which could tell us precisely where we stood between books. (Remember, in those days, an Arbitron sweep occurred just twice a year.for four weeks in April/May, and for 4 weeks in October/November.)

So, my wife made 150 random calls a week (from our ?kitchen table research department?) which gave us a clear indication of where we stood relative to our direct competition. And in that way, whenever Chicken Little sensed from his anecdotal ?surveys of one? that we were going to hell in a hand basket, I could whip out last week.s Radio Index research and show him, ?No, see, we actually increased from 10.7 to 11.2%?. That shut him up rather smartly. That was where we stumbled upon the concept of ?Preference? listeners (or P1 partisans as they later became known) in relationship to various different levels of cumulative listening.

That was where I discovered a pocket of listeners that I called ?Invisible cume? (which was later also known as ?phantom? cume or ?long-term? cume) and how it could be awakened quickly to produce quick ratings gains.

All of which, brings us around to music research. Every week, like most radio stations, we dutifully called all the major record stores in the Phoenix area to get their report about what we assumed was their Top 20 selling songs. I was finally able to allocate a few dollars to do the record store project I?d pitched to Paul Drew a couple of years earlier. So, we targeted the largest record store in Phoenix, and got permission from them to ask more in-depth questions of record buyers (about what other songs they?d buy if they could afford it, and any songs they didn?t like or were getting tired of.)

That is where we had a major mind-boggling breakthrough. For several weeks, we spent Saturdays there (figuring Saturday would be their biggest sales day, so we?d be able to talk with the most people.) Mind you, this was the largest store in Phoenix. To our amazement, they only made something like 15 or 20 actual sales of singles on a typical Saturday, their busiest day! In other words, we suddenly realized that, when they were giving us their Tuesday sales reports, they were basically just reading back our chart to us?there?s no way they could have actually calculated a true ?Top 20? from their actual sales. And of course, the smaller stores had to be totally fabricating their reports?outright lying to us. We also realized that from interviewing just the 15 or 20 so called ?record buyers? we talked with, we didn?t really have much to tabulate in terms of what else they would have bought or what they didn?t like. That was a real eye opener?realizing that record sales reports we were getting every Tuesday were essentially bogus (and not worth the time we spent doing them.) The store clerks were just telling us what we wanted to hear or probably, what their personal favorites were. It was totally worthless as quote, unquote ?research?.

But I wasn?t ready to give up on the idea of ?Q-scoring? our music. I felt that the premise of asking people what they liked and didn?t like was still a very sound one. I just realized that it wasn?t going to happen in a record store; it had to be taken to another level.

Since we were already doing in house callouts very successfully to track our week-by-week ratings progress, it was a logical extension to apply that callout methodology to the music research. Zero in on ?our? listeners, both P1s and cumers, and ask for their opinions about every song. Only problem: again, we didn?t have any budget for such a new and untried concept. So the idea sat on the shelf awhile.

I moved on to become PD of KTSA in San Antonio, where we further perfected the Radio Index weekly ratings tracking concept. Then a year later, Gary Stevens, who was GM of Doubleday?s KRIZ, lured me back to Phoenix as PD and Morning Man with the promise that he would allocate a small budget for the callout music research project. This was late 1973.

Another big breakthrough for the system came when I decided to clone the diary keeper mentality for the music research project, which at that time I was calling MARS (Mass Acceptance Research Study.) After considering various methods, my reasoning was that, since we were in a diary measured world (the way Arbitron measured radio listening), wouldn?t it be great if we could find out what songs diary keepers, as opposed to mere ?listeners,? liked or didn?t like? It was easy enough to do. We were already making the Radio Index tracking calls, we just isolated ?our? core listeners and, just like Arbitron, asked them if we could send them a ?music diary?. It was a simple list of about 45 songs and we asked the respondent to rate each song in their diary, using a 1 to 7 semantic differential scale of fixed values. Our scale had two levels of positives, two levels of negatives, one level of neutrality, one value to reflect unfamiliarity, and one that pinpointed formerly ?liked or loved? songs that they were getting ?tired of?. Basically, there was no other way that a listener could feel about a song.

Later, we discovered this same scale of positive familiarity could be applied to measuring the relative popularity of other things too: like DJs, contests, programming features, artist images, and a whole lot more.

Now, the music diary was actually quite an efficient way of checking lots of songs, but to speed up the turnaround, we would call the respondents back to retrieve their answers. We figured that when we?re dealing with a perishable product like cutting edge ?current? music, you don?t want to waste a week of a song?s life cycle just waiting for a respondent to get around to mailing back their diary. Doing it this way, with ?callout? retrieval or to be more precise, ?call-back? retrieval, our response rates increased because our calls also served as ?reminder? calls.