inthebag said:
I believe the fact that's being overlooked here is that music research can, in some instances, be a useful tool, but it is not the absolute last word some programmers portray it as.
In most auditorium music tests, 450 to 600 songs are tested with a goal of assembling a playlist of about 350. That means in any given era of music, several thousand songs will not even make it to the test. Therefore, you'll never know if those songs are hits or stiffs.
No matter how hard you work to qualify the participants in the study, you can never be absolutely sure you have an honest representation of your audience, or the market at large. Most studies also overload the test with the station's P1 listeners, so you're actually preaching to the choir when you should be looking for ways to make a few converts.
The auditorium test is a very artificial enviroment. Listeners brought in for these tests know they are part of a research study and a large majority of them give answers they think are proper for such a study, rather than what they really feel and think. Even peer pressure from the people around them affects their answers.
Add to that, the questions asked in such a study are incredibly critical, because if you ask the wrong question, you're going to get the wrong answer. In a test I was involved with in Denver some years ago, "Old Time Rock And Roll" hit a passion score that was easily one and a half times bigger than any other song tested, however that song also scored a burn factor of 98%. In other words, a lot of people liked the song, but they were tired of hearing it. If the burn factor had not been a part of that study (and it isn't in a lot of studies) the results would have been greatly in error. In other words, garbage in, garbage out, or as Mark Twain once said; "There are three kinds of liars...liars, damned liars and statistics".
A station's music research strategy must allow for a small dose of reality to come in. As an example, a few years ago, Paul McCartney sold out two nights at the Name-Of-The-Week Arena in Sunrise in less than three days. At that time, no station in the Miami/Fort Lauderdale metro was playing McCartney because he didn't test well. Pardon me, but the fact that about 40,000 people shelled out an average of $50 each to see the man, and a few thousand others would have bought tickets if they could have, says they might want to hear him on the radio every now and then.
No, I am not totally anti-research. I simply believe research will only give you part of the picture, and a possibly skewed picture as well. That's where a talented, skilled PD or MD's gut comes in. And, as it was so well stated earlier, music is only a part of a winning station. Execution, personality, stationality, imaging, street visibility, promotional stance and a number of other factors determine a station's success, no matter what the music format may be.
I often think back to what a PD I worked for in the 70's always said..."If we could write a book on exactly how to program a winning station without any guesswork, we'd all be gazillionaires, because whenever you think you know exactly what the listeners want, they'll be more than happy to show you just how wrong you can be". Unfortunately, some programmers today think they wrote that book, and they continue to recite their "absolute truths" as audience numbers decline.
By the way David, you said you programmed a classic rocker in a market of 18 million back in 2001. Currently, Arbitron shows New York (the #1 market) with a population of 15,332,000. So where was this market of 18 mil?
I have done AMTs with as few as 250 songs (CHR derivative) and as many as 1,600. There is no set length, and no set goal of acceptable songs. However, once you test several times, you start realizing which songs never test, which test after resting, and which always perform. Each format, even in far-apart markets, has the same core number of songs, and what makes each market different is the amount of play and familiarity created by the entire competitive array over all the years the songs have been available.
Every time we test, we put in "what if" songs. Hundreds and hundreds of them get tried just on the chanc they may work. As I said, when you see the same song score a 70% rejection (hate and dislike) every time, you stop testing it... but we constantly look for songs that might fit and might test.
At music tests, listeners are told that they can affect the music they hear on the radio. They are empowered to vote the way they feel, not the way th eperson next to them feels, not the way thier wife or husband or brother or friend feels. They truly get the fact that they are indicating how much or how little they owuld like to hear the song on the radio today.
Some tests score subjective values. I prefer those that simply score, 1 to 100, how much you would like to hear the song. And then follow up with a test at least every 6 months, more in bigger markets... I have some that test 6 times a year!
Ah, the concerts vs. radio airplay issue. Talking about Miami, we have a market of nearly 4 million. 40 thousand at a concert is 1%. But more than that, it is less than a 15th of the cume of the top cuming station in the market and less than a thenth of the cume of, let's say, Y-100. We have no way of knowing which songs, if any, the bulk of an audience may want to hear. In fact, we do not know if any of the attendees expects or even wants to hear on the radio. I have been to plenty of concerts where I had no interest in hearing hte songs on the radio, but liket the concert environment considerably. This argument also fits the "they are playing it at all the clubs" comment, too. Certainly, Cd sales, concerts, clubs, etc., are part of the things one should know about. But, mostly, you want to know about how to improve the radio listening experience.
There are many other factors. What Bill Tanner calls "the glue" that holds a station together, including talent, degree of personality, imaging, signal, processing, promotion, the competitive array, etc. But you can not have a music based station with music that is not right.
If we give a good carpenter knotty or rotten or warped wood, he will have ahard time building our bookcase. If you give me the tools and good wood, I will have a hard time as I am not a carpenter. So a good PD with good research is the proper combination... the ability to make every hour flow, not just take the songs as scheduled. The ability to know what songs may require dayparting or which, even if they test, don't belong on the station. And the ability to put the talent and other elements into the mix. But if you tell a PD that they must do this, but play (as suggested in another post) 1600 songs, it will fail. Nearly all of us have had one learning experience where we thought "variety" means "many" rather than "good" and we don't do that again!
The classic rocker was Emmis' Mega 98.3 in Buenos Aires, Argentina, the second largest city in the Hemisphere. The market, at any one time, has about 100 radio stations, including full ocverage (Mega was 200,000 watts) and neighborhood stations and is among the most competitive markets in the world. And not only was the station classic rock, it only played music by Argentine artists. The full story is here:
http://www.davidgleason.com/1999-Mega.htm