Tuesday, July 7, 2009

What We Play

MusicMaster Oldies drives 'real' radio programmers crazy. For the past 40 years or so, the buzzword in radio has been 'tight' playlists. The idea is to capture a loyal audience by playing only the most popular songs within a carefully selected and filtered genre. Radio became Top 40 Radio, then Contemporary Hit Radio (CHR). The 'format' went from Music to Hits and Country, and then split again and again into things like Urban, Alternative, Classical, Triple A, Modern Country, Jazz, Classic Rock, R&B, Classic Hits, 60's, 70's, 80's, 90's, and many more. Some types of music didn't fit into any of these formats and have gone nearly extinct on the radio. Some music was deemed too 'dated' to keep playing, so those songs disappeared from the airwaves as well. Even more songs were removed because they were determined to be 'Not Politically Correct.' Still more went away because of audience research. In fact, audience research has probably caused the disappearance of more music than all the other factors combined.

Basically, a radio station that wanted to build a large audience would regularly 'test' the music they play with a sample of their target audience. This could be done over the phone (Callout Research), or by inviting a group of listeners into a meeting place (Auditorium Research), and more recently by letting them voice their opinions online, either via a website or email (Internet Research). Research worked very well for radio. The better research the station did, the bigger share of the audience they would get, provided they adjusted their playlist accordingly. If you were planning to put a new station on the air in a certain town, you'd want to do research first. Let's say you determined that no oldies or jazz stations existed in that city. You'd start by doing research to see which format, Oldies or Jazz, had more potential listeners in your town. Let's say Oldies won this battle. Now, you'd determine what age group you'd like to attract to your new radio station. One logical way to do this is through some basic psychology. If you are planning to play Oldies, let's say from the 1970's, the 'center' of your format would be music from around 1975. The theory that drives this logic is that the majority of people form their 'favorite' songs and music styles during their High School years. Those who graduated from High School in 1975 were 18 years old back then. Fast forward to 2009, 34 years later, and you determine that your new 70's Oldies station will most appeal to people who are 34 + 18 = 52 years old. If you think you can make money with a radio station that appeals to 52 year olds, you can go ahead and build your 70's Oldies station. After all, making money is what business is all about, and radio IS a business. There's NOTHING wrong with that!

Now you put together a list of songs that you might want to play. Professional radio programmers have learned to pretty much ignore their own personal favorites. They might go back to the sales records for 1970-1979 and see which records were the biggest hits. From that list, they'll probably remove songs that they feel are dated, no longer politically correct, or too far away from the mainstream style. Novelty records are usually removed, too. Once you're heard a 'joke' it's no longer funny, and the same goes for Joke records.

Now you've got all these records from the 1970's that you might play. So you're ready to invite a bunch of 49 to 55 year old men and women into an auditorium so you can get their opinion. The way these people give their opinions can vary, but the basic method of testing music is pretty standard. First of all, you don't play every song for them. There's no time for that. You also don't have time to play entire songs. You may hold their attention for a couple hours, but you won't be able to keep them prisoner there for days or weeks! So, you have to remove a whole bunch of songs from your 'test' list first. Which songs get removed? Well, maybe it makes sense to take out the lesser hits. The songs that didn't sell well, or get played on the radio as often back in the 1970's, get removed from the list. Now you record 'hooks' for each song, usually several seconds from the middle. Ideally, this hook should be the part of the song that gets stuck in your head after you've heard it! Selecting the right part of the song to play in a hook has become an art form among radio folks. The hook may, or may not, contain the part of the lyrics that contain the title of the song. (Homework: Go listen to some songs you like and see if you can figure out which seven seconds you'd play out of that song if you were testing it.)

The audience reaction may be measured several different ways. For one thing, you could just ask them to rate the songs on a scale, like 1 to 5, where 1 is a song you hate and 5 is a song you love. Some programmers prefer a 'binary' approach that dials in on the listener's preference by asking yes/no questions. After playing the hook, they might first ask, "Do you know this song?" Then they'd ask something like, "Would you like to hear this song more often or less often on the radio?" If they answer that they'd like to hear it more often, the next question might be, "Is that because you just like it, or is this one of your favorite songs?" If they say they want to hear it less often, the follow up question might be, "Is that because you're tired of hearing this song, or because you never liked it." These responses allow you to extract more than just a Like or Dislike response. You also get to know things like Familiarity (how many people know the song?) and Burn (how many people are tired of hearing the song?). Another test method involves using a dial, just like a volume control on the listener's radio. The audience is instructed to turn the dial up if they hear a song they like, and down when they hear one they don't like. Pretty cool, eh? This method has some very interesting aspects and has also become somewhat of an art form. Researchers might pay attention to how far the listener turns the volume up, or how quickly they turn it up, or maybe at which point in the hook they turned it up! Each of these are clues to how much the listener likes the song, which is basically the thing you're trying to find out by doing this research.

When the test is over, you compile all the scores. Now you eliminate the songs that had a high Dislike score overall. Most programmers also remove the songs that didn't score well on Familiarity. The idea is that EVERY song you play should be one that MOST listeners know and like very much. This is based on the idea that, when trying to appeal to a mass audience, you don't want to play any song that a bunch of people dislike. If they hear a song they don't like, or don't know, they may tune to another station. Once they leave your station, it's VERY hard to get them back. Certainly nothing you do on your station will get them back -- because they're no longer there to hear it.

There were at least 50,000 different songs in the 1970's that people liked enough to buy. Out of those, the auditorium audience may have heard hooks for about 500 of these, or about one percent. Out of that 500, only about 250 of them may actually make it on the air on the new 70's Oldies station, with the others being rejected because they didn't do well in the research. If you've ever wondered why your favorite Oldies station doesn't play some of your personal favorites, maybe this research business explains it. Your favorite probably didn't 'test' well.

The research model may be biased. It may be based on some bad assumptions. For example, what if some of those songs that didn't sell well back in the 70's would test VERY well today? After all, times have changed. The audience has aged and matured. They've been exposed to a lot more different music since the 70's. Even though they still have a special place in their hearts for the hits of their High School years, there are probably songs they didn't like back then, but would totally love today. The exposure they've had to 40 more years worth of newer music has influenced their preference when it comes to music. That's logical, right? Many radio stations will go back later and conduct more auditorium tests later using another group of songs from the target era, just so they can find more favorites and add them into the playlist. Unfortunately, this sometimes means removing songs from the playlist to make 'room' for the newly discovered gems. Radio stations also have to think about how often the listeners hear each song. Songs that everyone likes have to be played more frequently than songs that appeal to fewer people, or so the theory goes.

Another bias might come in the automatic rejection of songs that are (subjectively) considered dated, politically incorrect, or outside the mainstream, and are therefore never exposed to the test audience for measurement. What if the audience would have REALLY loved some of those songs? If you remove them from the test before you start, you'll never know.

The ultimate playlist on your new 70's station might also have other problems. What if there are songs that test VERY well when the audience hears them for the first time after not hearing them for decades -- but that same audience gets sick of hearing them after the second or third time? Testing them by playing the hook only ONCE doesn't tell you anything about the 'life' that song may have. Some songs are going to make the audience have an 'Oh Wow!" reaction, for sure. But some of those songs are going to continue to be loved no matter how many times you play them. Others are going to quickly 'burn' and cause your audience to tune to another station. Radio stations learned that they needed to continue doing research on an ongoing basis AFTER putting their new station on the air just so they could identify and eliminate these troublesome songs from their playlist.

There are many other factors that can bias the research results. In fact, there are too many of them to cover in this blog entry. Just to give you an idea, what if you didn't pick the right people for the audience test? What if you played the hooks in the wrong order? Maybe people give a lower score to songs they really like if the previous song they heard was a huge favorite. Maybe people give a higher score to a song they don't particularly like if the last three songs they heard were real clunkers. You get the idea, right? If you give all this some thought, you can probably think of a whole bunch of other factors that might influence the results.

What I'm doing with MusicMaster Oldies is NOT based on research! Basically, I'm playing EVERY song that people liked enough to buy. Every single song that ever showed up on the sales charts (Billboard, Cashbox, and others) is on the air on MusicMaster Oldies. Songs are not removed because they are dated, politically incorrect, or for any other reason.

On top of those nationwide (USA) hits, other 'Regional Hits' have been added. These are songs that showed up on local radio station playlists but did not make it onto the national charts. Some of the charts I use for this include the WLS Silver Dollar Survey (Chicago), the CHUM Charts (Toronto, Ontario, Canada), and MANY others. In addition to these, I've also included EVERY song that ever made it onto the "Bubbling Under" chart. These are songs that sold enough to get 'honorable mention' by Billboard Magazine, but never sold enough to make the Top 100.

I've also been adding songs that have appeared on the sales charts from other countries around the world. You'll hear songs that were 'hits' in England, Germany, France, Italy, Spain, Mexico, Singapore, Japan, Australia, New Zealand, and MANY more countries. Some of these songs were never published or played in the USA, but they were certainly influenced their contemporaries. The Beatles didn't just influence a bunch of midwestern garage rock bands in the USA -- they influenced musicians in every corner of the planet -- and they STILL do!

MusicMaster Oldies also contains a whole bunch of songs that NEVER made it onto a record sales chart or radio station playlist. These are songs that could have been hits, and maybe should have been hits, but for whatever reason they were overlooked. How could that happen? Well, don't look now, but it's STILL happening today! Many artists don't get the credit they deserve because they haven't been 'discovered' yet. They may not have been promoted well enough. There are so many factors that go into this that it would be impossible to list them all here. For example, some artists were rejected by the record companies or radio programmers because they 'sounded too much' like some other artist who was already making hit records. In the 50's, if you sounded too much like Elvis, it was a problem. In the 60's, if you sounded too much like the Beatles or the Rolling Stones, the record companies may not sign up your band, or the radio stations might not play your records. I'll be the first to admit that many of the songs I've added that were never 'hits' are the result of my personal, subjective opinion. But, I'm also guided and influenced by the collective tastes of others who, like me, have rediscovered the music of the 50's and 60's and are now busy restoring rare old vinyl records, pressing them on CD's and selling them to 'new fans' all over the world.

There are also songs on MusicMaster Oldies that were never hits back then, but have since found huge fans in strange and interesting places. One such group of songs is known as Northern Soul. These are usually uptempo R&B non-hit songs from the USA that are great 'club' records. They're great for dancing! But, many of them were never played on the radio. The owners of clubs in northern England came over to the USA and started digging through piles and piles of discarded records in bargain bins and resale shops searching frantically for buried treasure they could play in their clubs back home. The regulars at these clubs loved the songs. They had the 'sound' of the 60's, but they were 'new' because they'd never heard them before. It was like going back in time to discover more great songs and artists from those years.

MusicMaster Oldies is currently playing 109,766 songs. More are being added every week, sometimes as many as 200 a week. Coming up, I'll explain how these are scheduled by MusicMaster, the music scheduling system, to create the unique and ever-changing playlist you hear on MusicMaster Oldies. In the meantime, please enjoy the music!


  1. I appreciate the background on radio station playlist selection! I suspect that more often than we realize, the songs that have a "oh wow" factor right away do end up getting burned out more quickly when they're only playing those 40 songs. Consider what would happen if oldies stations did what the contemporary stations do and changed the entire playlist of 40 every so often. It would be as though we were actually back in the seventies, and the catalogue of hits was being slowly replaced as new songs came in.

    The flip side of that is: do I want to hear "seventies music" or do I simply want to be able to tune in and have a chance of hearing my favorite 70s song? If the latter were the case, I wouldn't want the station to change up the playlist, nor would I want them to play such a wide variety of songs that my song only came up every few days. So there's a completely different paradigm here. Each seem legitimate, but the stations serve different purposes. One is like a jukebox, the other is a museum.

    I get the sense that MusicMaster Oldies is more of a museum, catering to those who want to hear fifties and sixties music, rather than one or a dozen particular songs. And I love it! You're finding music to play from the same selection that radio programmers had back in "the days": a hundred thousand songs to choose from, both good and bad, popular and obscure, cheesy and enduring. If there's a station that can conquer the oxymoron of "new oldie" then I want to hear it!

  2. One reason why radio is not popular anymore as it used to be is toouch boring repetition on too tight playists. I appreciate the variety and the surrise factor, hearing stuuf ahve never heard or not heard or a long time...therest I cn do on my Ipod filled with 14.000 songs.