How "current" is the Nielsen reports. One would think if there is a "dropoff" in a 15 min. block of P1's you could go back and see what was played that quarter hour. Check the drop-off quarter-hours and if there is a song(s) that is common to the bad quarter hours, pull or at least retest the questionable song(s).
-
Get involved.
We want your input!
Apply for Membership and join the conversations about everything related to broadcasting.
After we receive your registration, a moderator will review it. After your registration is approved, you will be permitted to post.
If you use a disposable or false email address, your registration will be rejected.
After your membership is approved, please take a minute to tell us a little bit about yourself.
https://www.radiodiscussions.com/forums/introduce-yourself.1088/
Thanks in advance and have fun!
RadioDiscussions Administrators
CBS-FM Adjusting Playlist
- Thread starter coreynj2461
- Start date
I must be getting old. I would be changing the station if I heard any of these songs pop up on WCBS.
K.M. Richards
Program Director, The Eighties Channel™
Subscribing stations in PPM markets can see minute-to-minute ratings, if they so choose.
K.M. Richards
Program Director, The Eighties Channel™
A, you and I both have access to Mediabase. We could (if we cared enough to, which I personally don't) pull up WCBS-FM's daily monitor and see.
In addition to the '00s titles, CBS-FM's scheduling and title mix have become noticeably more uptempo. In recent years, CBS-FM had often sounded as if it was trying to wrest the soft AC franchise from Lite, but I'm now hearing much less Chicago, Heart, and Rod Stewart and more Guns n' Roses and Aerosmith. Some of this is surely a function of format evolution and research; some may be a competitive response to Q104.3.
To my ears, CBS-FM sounds significantly better than it did during the pandemic. The imaging has markedly improved, incorporating listener drops and timely/topical flourishes. The morning show does a nice job engaging listeners and makes a legitimate effort to be local on- and off-air, something plenty of stations/shows stopped bothering to do years ago. All of the weekday jocks sound great, and I've even heard cross-talks between Foxx & Annie and Race Taylor, a formatic element I hadn't heard on music radio in eons. To evoke Sean Ross, I get the sense from listening that there's a bit more going on than there was 18-24 months ago.
None of this is revolutionary -- some of the music and imaging tactics are straight from the K-Earth/Audacy Classic Hits playbook. I suppose there's also a risk that some of the tweaks to be more foreground could cost workplace meters, or that they become too texturally wide. However, Lite FM certainly hasn't suffered for playing Doja Cat and Chris Brown alongside Richard Marx and Berlin, nor has K-Earth suffered in LA for playing AC/DC at 9am on Monday morning. For now, I'm glad the station sounds a bit more dynamic and think the collective adjustments have improved the station.
To my ears, CBS-FM sounds significantly better than it did during the pandemic. The imaging has markedly improved, incorporating listener drops and timely/topical flourishes. The morning show does a nice job engaging listeners and makes a legitimate effort to be local on- and off-air, something plenty of stations/shows stopped bothering to do years ago. All of the weekday jocks sound great, and I've even heard cross-talks between Foxx & Annie and Race Taylor, a formatic element I hadn't heard on music radio in eons. To evoke Sean Ross, I get the sense from listening that there's a bit more going on than there was 18-24 months ago.
None of this is revolutionary -- some of the music and imaging tactics are straight from the K-Earth/Audacy Classic Hits playbook. I suppose there's also a risk that some of the tweaks to be more foreground could cost workplace meters, or that they become too texturally wide. However, Lite FM certainly hasn't suffered for playing Doja Cat and Chris Brown alongside Richard Marx and Berlin, nor has K-Earth suffered in LA for playing AC/DC at 9am on Monday morning. For now, I'm glad the station sounds a bit more dynamic and think the collective adjustments have improved the station.
The slower the tempo, the older the demo.
Looking at minute-by-minute data is not actionable, the sample size is not large enough.
The ratings you see published are for an entire four week period. And many buyers and programmers look at three month periods to smooth out wobbles.
If you were to look at one station at the time of one specific song play, there may only be a handful of meters tuned in, not enough to make any valid assertions. Demographic information is also not available at such a granular level (is that a 20 year old or 70 year old tuning out?)
A product called "Mscore" does take the average reactions of meters to specific songs over the course of a week, but even that is not truly actionable. You don't know why a meter turns off a station (did they not like the song? or is that when they got out of the uber that had the station on?)
A song may have a negative score one week based on meter reactions, and then be through the roof the next week. Which week was the true representation? That's why programmers still rely on research where they can control the parameters of the sample.
There are too many variables for this to be totally accurate and that is why stations test their music independently. As Huff said, there is not enough actionable information.
The time in a quarter hour a song is played may affect results... such as the moment just before the hour or half hour when people get out of their car to go in to the workplace. And you don't know if tune-out was due to an early or late start of a stopset, the fact that this was the second song a listener did not like in a row, etc., etc. and etc.
But the biggest reason is that there are not enough meters "hearing" any particular station in a given quarter hour to be an adequate sample. A top station might have 8 to 12 meters at a given moment in time. To test currents, somewhere between 60 and 80 participants is thought to be the minimum for accuracy across gender, age and, sometimes, ethnicity.
Here is the math for an imaginary larger market: There are 2000 meters. Persons using radio during the whole day averages 6, meaning that out of the 2000 meters, only 120 are actually "hearing" a station. And a station with a 6 share across the day will have an average of... note this... 8 meters detecting them.
No. Don't confuse cume with average quarter hour listening. While there may be only 10 or 12 meters detecting a station in one quarter hour, remember that the average weekly listening to radio is about 6 hours. So each quarter hour, each hour, each day is made up of detections of many meters in total.
So a station with a share of 6 might have a cume of 20% to 25% of all listeners... in that same theoretical 2000 total meter sample, perhaps 350 to 500 meters may show listening to that station.
But what can affect ratings is the loss of a larger household that has several or all members cuming one station regularly. If replaced with a household that does not listen at all but meets all the other variables like ages and ethnicity, then one or more stations may be affected a bit.}
That is one of the issues where the sample size is as large as stations can afford to pay, even if a larger sample would reduce those wobbles.
And just like that, I'm not longer a "young adult." Because that's *MY* music (circa 2003). In fairness, I turn (sigh...) 40 next year.
I'm now just an adult who listens to "oldies." When will they start getting re-sung Reelworld jingles from the late 1990s, they way oldies stations used to get PAMS re-sings?
One would think if everything is going good then all of a sudden you take an audience hit for a quarter hour every now and then (for no logical reason) you would look at the songs (or what the announcer is doing) 8 minutes in the preceding quarter hour and that quarter hour and see what is happening?
I doubt a human could take the time to do this but a computer could do this quickly even detecting long periods of no music or no commercials where the on air talent is not so talented.
Thankfully nobody did this when I was on air!
I am thinking about possibly a Rap song that could be offensive to females gets into the CHR playlists, non county artist with a "hit" on country charts, or some kind bit the announcer does.
Again, the sample is so small and there are so many independent and non-music-related variables that no conclusion can really be made.
If you are #1 in San Antonio or any other "bottom tier" of the PPM markets, being #1 may also mean that you have 5 or 6 meters detecting you in any given quarter hour.
The final monthly ratings are not based on single quarter hours. They are based on 18 total days. So little blips average out. As Huff mentions, we can get Mscores for weekly plays, but this is still based on adding a whole lot of random plays, each of which is subject to other variables. We'd like to think that nobody is playing a stiff that is so bad that it can cause repeated tune-outs.
Remember, Nielsen does not do AM, or "Average Minute" ratings. They do "Average Quarter Hour". There is no minute by minute rating data because, to a certain extent, not every audio segment on a station give the PPM opportunity to "inject" its data bits.
How can you say it's small? 5 to 3 meters times 20 or 40 events is a decent sample size. (Hopefully you don't have that many)
You are taking 22 or 23 minutes (Quarter Hour plus preceding 2 or 3 songs). There might not be anything in common during the time periods but then again maybe your sample that the music was tested on is flawed and a stiff snuck in.
Basic Quality Control: If you have defects on a product even if it's less than 1 percent any QA manager worth his salary will examine the defects. Especially with AI's ability to quickly studying weak points in any product, defect analysis should done in a lot of business not just radio.
Good quality control (urban legion they started with the AT&T equality control handbook) won the Japanese auto manufacturers market share (along with good fuel mileage).
I know "we never done that" is prevalent in a lot of business. But I will bet a small cup of Burger King regular coffee if someone with the IT ability does this it could stop "tune out". Based on revenue, I am surprised someone in the top 5 markets hasn't done this. More I think about it maybe I am giving away something.
K.M. Richards
Program Director, The Eighties Channel™
You have obviously missed David's point, so allow me to attempt a rephrasing.
The flaw in your presumption is that the PPM tracks only intentional listening when it does not. The PPM detects any audio source which has been encoded and logs it because the wearer could hear that audio, even if not focused on it.
You could be a person wearing a PPM using a rideshare and if the driver has the radio on, the PPM will log that as listening until you get out of the car. Similarly, if you are driving yourself and listen to the radio, the PPM will stop detecting when you get to your destination and turn the engine off. Neither of those circumstances means that the tune out had to do with what song was playing at that moment.
If I spend ten minutes in a store which (illegally, but that's not the point) has the radio on for background music, the PPM will count that as my "listening" as well. Again, when I left the store the timing was not based on what was playing at the time.
So, even if to some degree it is possible to correlate PPM data against the playlist, that is not reliable data to make a determination of a song's relative popularity.
My point is that there are so many other variables beyond the song itself that even an amalgamation of many incidents is risk-ridden. And we don’t have a balanced sample by demographics.
In the case of a specific song, we need to know if any negativity is even in our target demo. And then, is it among younger and or older persons… or men or women… or by ethnicity.
Music test samples are professionally recruited. I have written “recruit specs” for over a thousand music tests, and can give you a simplified version:
First, we look for either our core listeners or a combination of them and certain overlapping and sharing competitors. We specify the age range of our core and the gender percentages. We require a certa in minimum number of hours of radio usage to the core station or the multiple stations to guarantee familiarity.
Sometimes, in competitive situations, we will play a montage of song hooks and only those scoring high on those that are like or similar to our format will qualify.
Participants are requalified when they take the test. And after the test, any that are outliers are eliminated as their results would taint the majority. It is unusual to find more than one or at most two outliers in a 100 person test.
You are complicating a simple process of finding out how much your listeners want to hear each song today.
We have been doing this for about 50 years now.
Only 18 days? Are certain days removed from the 28 day period?
Saturdays and Sundays, I'd imagine, as they're ignored in ad sales.
- Status