New York Metro Radio Ratings: June 2023

[MarkW]

I can think of two potential examples:

WLIT Chicago (when MyFM was dropped in favor of relaunching the Lite FM brand).

WFLC Miami (when CHR was dumped in favor of Rhythmic AC).

I suspect the number of unique songs played in the span of a week increased after these stations' respective changes.

BigA raises a good point; perhaps a more useful metric would be number of unique songs played in a four week span?

 

[TheBigA]

BigA raises a good point; perhaps a more useful metric would be number of unique songs played in a four week span?

The high cume and low rating tell me that there's a lot of audience turnover. Perhaps there are others having the same experience as the previous poster: They like the station, but encounter a song they don't like, and they leave. That's typical for stations with larger than average playlists.

 

[davideduardo]

I can think of two potential examples:

WLIT Chicago (when MyFM was dropped in favor of relaunching the Lite FM brand).

WFLC Miami (when CHR was dumped in favor of Rhythmic AC).

I suspect the number of unique songs played in the span of a week increased after these stations' respective changes.

But those are format changes. I was referring to stations that were not happy with ratings in their existing format, and which expanded the playlist to broaden appeal.

BigA raises a good point; perhaps a more useful metric would be number of unique songs played in a four week span?

That's easy to do with MediaBase.

Many stations with larger gold libraries have songs that either play less than once a week or which are cycled in an out. A typical example would be songs that play once or twice in each eligible daypart, and then rest for a period of time while others cycle in.

Typically, a the categories that have such options for songs will have a few songs enter and a couple leave each day. So, over a period of 4 weeks, there might be 50 song "positions" in a rotation category but over 100 songs that come and go, with at least half always on rest. Both Selector and MusicMaster have built-in systems that will do this automatically in accordance with the rest time and rotation each song is to have.

 

[TheBigA]

That's easy to do with MediaBase.

It's easy to compare this week with last week. To go back a couple weeks takes more effort. Or perhaps you do a split screen on the monitor.

 

[davideduardo]

A lot of stations replace songs every week. Not many replace 2/3rds of their list every week.

I've never seen "weekly" replacements. Every station that I know or have observed does the rest/rotation cycle on a daily basis, where a few songs cycle in and out every day. Some, with a desire to get greater variety perception, will have a song rest for a determined period once it has played once or twice in every daypart with a specific separation in hours for each play.

I prefer the later, where a rotation cycle is done and then a rest period of similar or greater length occurs. This system can also be used in "packets" of songs by an artist with "too many songs" to rotate different titles in and out of a slot for a particular artist.

 

[davideduardo]

It's easy to compare this week with last week. To go back a couple weeks takes more effort. Or perhaps you do a split screen on the monitor.

Or you create a longer period of measurement. Otherwise, it seems that the songs use weekly cycling, when the rest/return period may be based on total plays before resting or a fixed time period that is not a 7 day multiple.

Yes, it was easier in that other system that went away... BDS. Harder was when we had the totally week-based separate Billboard chart.

 

[TheBigA]

I've never seen "weekly" replacements. Every station that I know or have observed does the rest/rotation cycle on a daily basis,

I'm looking in the 7-day format. I can change it to daily and get those numbers too.

Or you create a longer period of measurement.

There you go. As I said, I look at the 7 day playlist.

 

[davideduardo]

There you go. As I said, I look at the 7 day playlist.

People in the music business seem to always look at weeks, not other periods of time. That matches the charts, sales reports, business plans, etc., while it does not match radio.

But most programmers look at rest periods that are not 7 day multiples. You want the songs to stagger, particularly on workdays, so that they change at intervals that dissect either calendar or Nielsen measurement weeks. That would mean making rest periods last either less than 7 days or more than 7, 14, 21 days.

You might even want songs that have rested reappear more on Thursday and Friday if you are in a diary market where you could give a perception of variety on the two key days of diary entries. In the PPM, you would want songs that return to have rested an even number of days so that you don't get odd day repeats (a week has an odd number of days) that can cause enhanced perception of repetition, particularly on weekends.

 

[TheBigA]

But most programmers look at rest periods that are not 7 day multiples.

The programmers are also able to compare the music with the daily and hourly ratings, so they can see if the music policy is what's causing tune-out.

My view is that music policy is not the problem with this station. I don't even blame the signal. My view is that the NY audience for this station still doesn't know it exists. They aren't dial scanners, they likely don't own radios, and they already have a place they go for this music. My suggestion would be for this station to do more community outreach. Put the station in front of the audience.

 

[MarkW]

It'd be interesting to know how the station stacks up versus Power and Hot in key demo categories.

In 6+, it seems Power has slumped since The Block came on the air.

 

[davideduardo]

The programmers are also able to compare the music with the daily and hourly ratings, so they can see if the music policy is what's causing tune-out.

Would a programmer do a music test with just 7 people?

In LA, somewhere between 6 and 12 meters will get you a #1 rating, depending on the daypart. Tune-out of 10% to 15% at any moment can be caused by one person who got to Ralph's and got out of the car. Using moment to moment behaviour of the PPM panel is like falling on your own knife; there is not enough data behind single day moment by moment listening to be anywhere near conclusive.

In smaller markets there are even fewer meters, so a station with a 5 share might be getting 4 to 6 meters in any given quarter hour.

The math in a market with 1000 meters: Average 6 AM to Midnight PUMM is about 6.0. So on average, there are 60 meters detecting actual listening. A station with a 5 share has 3 meters in an average quarter hour. Of course, it changes by daypart, as PUMM is higher in Midday, then in Afternoon, then in Morning and then in Evening. So a PM drive 5 share may actually mean 5 to 6 meters, and at night perhaps 2 of them.

My view is that music policy is not the problem with this station. I don't even blame the signal. My view is that the NY audience for this station still doesn't know it exists. They aren't dial scanners, they likely don't own radios, and they already have a place they go for this music. My suggestion would be for this station to do more community outreach. Put the station in front of the audience.

It's been reported that they were using outdoor and transit in the primary coverage areas. I don't know if this is a fact, but for NYC that is a very effective medium to build awareness but it only works if the product validates the test listen by those who see the ads.

In a non-US market where I have "involvement" there are loads of what they call there "digital billboards" which change image every 8 to 10 seconds. One of the group stations actually has a picture of every artist it plays and when its turn on the digital boards comes up, it shows the artist and, roughly, "Right Now On (station name) (Station Frequency)" The ad is a still, not a piece of a video, as running the video would have lots of shots where the artist is not seen.

One of the staff members did a video of driving by several of the boards when our ad was on, and it really gave me chills to see it.

 

[TheBigA]

Would a programmer do a music test with just 7 people?

They supposedly have already done the music testing, and the result is a 700 song playlist with comments like this:

I change the station when I hear Ashanti and Mariah Carey with ODB.

It's one thing to test songs, and another to test those songs in context, in real time, next to other songs.

It's been reported that they were using outdoor and transit in the primary coverage areas.

There is no replacement for the personal connection. How do you get a million listeners? You shake a million hands.

The digital billboard would be more effective if it actually played the station so people could hear what they're doing.

 

[davideduardo]

They supposedly have already done the music testing, and the result is a 700 song playlist with comments like this:

What I am saying is that to use moment by moment Nielsen PPM data is an "immensely small" sample and totally irrelevant and unreliable.

It's one thing to test songs, and another to test those songs in context, in real time, next to other songs.

But the PPM data is based, for most stations and dayparts less people than I have fingers and is subject to multiple variables such as tune-outs based on factors totally unrelated to programming.

There is no replacement for the personal connection. How do you get a million listeners? You shake a million hands.

Not always. Sometimes if the effort is put in the programming, the listeners come. I had a station with a cume of over 5 million, and we did very little street except for a few music events.

And if the target is under 35, there is a real, measured resistance to commercialized personal contact... like older people's aversion to used car salesmen but heightened by general commercial skepticism.

The digital billboard would be more effective if it actually played the station so people could hear what they're doing.

The average view time for an electronic billboard in the market I referenced is around 5 seconds. The minumum length for a hook in a music test is about 8 seconds, so those able to hear the billboard would not benefit from the sound. In any case, the audience for a billboard is vehicles in traffic, mostly on highways where viewers are in closed cars or public transport and would not hear anything (even if there were speakers on the board and regulations / laws allowed such loud noise).

I can't think of a single place one of those boards is located where audio would be allowed.

 

[TheBigA]

What I am saying is that to use moment by moment Nielsen PPM data is an "immensely small" sample and totally irrelevant and unreliable.

Don't focus on moment by moment, but rather the overall trend. If Mariah Carey gets the same results over and over, perhaps there's a problem.

Not always. Sometimes if the effort is put in the programming, the listeners come.

That may have worked 10-15 years ago. But fewer and fewer people have radios and they aren't sampling new stations the way they once did. You need to get to them where they are, and put your product in front of them. It's getting harder and harder to launch new stations.

 

[davideduardo]

Don't focus on moment by moment, but rather the overall trend. If Mariah Carey gets the same results over and over, perhaps there's a problem.

The problem is that to accumulate sufficient data there is going to be a considerable passage of time, which introduces an additional variable. Trying to track program elements via the PPM "instant" results has, to begin with, many lifestyle variables that have nothing to do with programming.

That may have worked 10-15 years ago. But fewer and fewer people have radios and they aren't sampling new stations the way they once did. You need to get to them where they are, and put your product in front of them. It's getting harder and harder to launch new stations.

It works today if the promotion is focused on content. In the specific case I mentioned, in another country, the station has a URL that is just the name and the dot and the country code. Something like mega.cl (an illustration, not the real one). It works, and in a nation with a higher percentage of smartphones than the US.

 

[TheBigA]

The problem is that to accumulate sufficient data there is going to be a considerable passage of time

They have to figure it out. Use all the tools available. Don't look for reasons to ignore information.

It works today if the promotion is focused on content.

The content is already available.

 

[Barry]

Is it quite possible that with 4 urban stations, this market is over saturated? Kiss 98.7, which was a much more established station than the Block was taken off the air about 10 years ago, leaving 3. 103.9 was a flop.
There were 4 Spanish language FM’s, till Univision sold 92.7. There are 3 sports stations. Other formats have 2 or maybe 3 stations.

 

[TheBigA]

Is it quite possible that with 4 urban stations, this market is over saturated? Kiss 98.7,

I think that was more of an Emmis thing, trying to lower their debt load, taking sure money from ESPN. They did a similar thing with their LA station. Since then, they've mostly left the radio business.

 

[Weiserguy]

My view is that music policy is not the problem with this station. I don't even blame the signal. My view is that the NY audience for this station still doesn't know it exists.

Sorry, that's crap. This station has been there for, approaching, two years. The absolute best promotion is (and has always been) word of mouth. And the word of mouth is that The Block sucks, or else it would have more listeners in the core areas it's targeting. It's not the signal, unless you're targeting White Plains, or Hempstead/Uniondale, or Stamford. I used to be able to pull 94.7 in from Valley Stream on a cheapo table radio. (I just never wanted to, because "Family Radio" was like nails on a blackboard to me, but their signal was superior to Jersey signal I did want to tune in.) So it should not be a problem in the 5 boroughs or close-in suburbs, or anywhere in North Jersey.

It's the music. It's the talent. It's the format. Most importantly, it's being #4 in a four-competitor slice of the market. It's a less-than-elegant solution to a nonexistent problem.

In a non-US market where I have "involvement" there are loads of what they call there "digital billboards" which change image every 8 to 10 seconds. One of the group stations actually has a picture of every artist it plays and when its turn on the digital boards comes up, it shows the artist and, roughly, "Right Now On (station name) (Station Frequency)" The ad is a still, not a piece of a video, as running the video would have lots of shots where the artist is not seen.

The average view time for an electronic billboard in the market I referenced is around 5 seconds.

Totally anecdotally: those e-billboards tend to get put in the wrong place. Near a busy highway, with cars trying to whiz by at 60, 65, 70 mph ain't the right place, unless your goal is to cause accidents. (The one near where I live, just before an exit off a major freeway, has done that more than a few times. Too bright, especially at night.) Unless the e-billboard is positioned next to the Midtown Tunnel toll plaza (or equivalent always-tied-up location), it's going to be ignored by drivers trying to arrive in one piece, and looking out for other drivers that aren't. (And if you're fixing your eyes on one of those billboards, you, by definition, aren't.)

The math in a market with 1000 meters: Average 6 AM to Midnight PUMM is about 6.0. So on average, there are 60 meters detecting actual listening. A station with a 5 share has 3 meters in an average quarter hour.

Which, for the application, is way too few samples. Probably by an order of magnitude. Sampling radio listening, which is an inherently mobile, interstitial (i.e., filling in the gaps between paying attention to more important things) style of listening, should not be being sampled the same way TV is (where you're in a chair or on a couch, paying attention to a program/stream/commercial or trying to decide what next to watch). TV is cognitively foreground most of the time, radio largely background. (With the exception of inherently foreground formats: news, sports, talk, where it helps to have your brain in at least first gear.)

Applying the statistical model for TV to radio is bastardizing the statistical mathematics. I wouldn't be surprised if the entire foundation of measuring listenership via PPM is flawed. I've read the rationales for doing it the way it's being done, but that does not mean the results are true and accurate.

Excuse me now, I need to chase the kids off the lawn.

 

[TheBigA]

Sorry, that's crap. This station has been there for, approaching, two years. The absolute best promotion is (and has always been) word of mouth.

That's 20th century talk. How many people are talking about WBLS? Or any radio station.

Outreach is the word. Go where the audience is. Engage.