First, there is the sample size of all people which is what Nielsen uses. Those samples are around one person out of every thousand for each survey.You're quoting ratings and stats as if the medium is on an upswing or even in its heyday. In my 41 years no one has ever asked me or anyone I know which radio stations I'm listening to or what I'd like to hear. The number of people who ever get a radio survey (and can be bothered to respond to it accurately - remember, what people say they want and what they actually want are two different things) seems to be a very small sample size.
Then there are research projects that a station does within its own target audience. A big perceptual in a market like NY or LA might cover 500 people, but all would be selected as liking a certain kind or music or format.
A music test might be around 100 people, all listeners to the format. So your chances of being invited to participate are very minimal. In the business we have done studies to see at what point a larger sample just gets more of the same results. What I just described is the product of that kind of study.
That is an old system, not usually employed for any significant period of time.It's pretty clear they're looking for quick ratings over maximizing future listenership.
- Play a concise heavy rotation playlist to give new listeners the vibe of the new station
If there are no more songs that get good listener reaction the library can't expand.
- Get some promising initial ratings
- Fail to adequately expand and freshen playlist
Not a proven fact. Most listeners want to hear "their favorites every time they listen".
- Listeners get bored and go elsewhere
Before that, a station does research.
- Ratings lag - flip the format!
Not all new products work in any business. Leader Proctor & Gamble has about half of its new, researched and tested products be discontinued after a year or two.
- Repeat until the carcass turns into a talk or religious station
Charts are not the basis for airplay. The question in a music test is "how much would you like to hear that song on the radio today?" Lots of songs that charted in the past are kisses of death today.I'm not asking for deep tracks, just a wider variety format of multiple genres and decades that were somewhat high on the charts. This could easily produce a 500 song playlist without a single unfamiliar track.