Lotus

[davideduardo]

Believe what you want to believe. There are listeners who are not getting counted, and that costs us money.

It does not matter if all listeners are not counted. The PPM is a panel, and between being theoretically 100% representative of the universe, it has a technique to replicated earbud listening based on a mathematical model of all other listening created with an actual sample of earbud using listeners representing all the different stratification variables.

Neither the diary survey (where earbud listening is registered) or the PPM (where earbud listening is mathematically replicated) counts all listeners. They both count, either as a panel or a random probability sample, a selection of listeners and project them into the intended universe.

This is exactly the same process as the Census Bureau uses for its annual sample-based estimates of US population, right down to individual towns and rural areas; in this case as well not all people are counted but just a sample.

 

[TheBigA]

It does not matter if all listeners are not counted.

You misunderstand what I'm saying. The Nielsen streaming data and the actual user streaming data don't match. Nielsen is undercounting radio streaming. That was the point of Pierre's article. I don't know or care why that is, but apparently it's true.

 

[davideduardo]

You misunderstand what I'm saying. The Nielsen streaming data and the actual user streaming data don't match. Nielsen is undercounting radio streaming. That was the point of Pierre's article. I don't know or care why that is, but apparently it's true.

Remember what Pierre's job is. He did just as good a hype job at Arbitron as he does now at his radio employer.

The main reason the data does not match is that both the diary and PPM measure actual listening by the participant. Streams are often left on with nobody listening.

Example: when we go out, we have Alexa stream so that the dogs have human company. If we are away for a whole day, we leave streams on rather loud to make thieves think someone may be home.

When people stream, they leave the stream on if they go to the kitchen or a meeting or whatever. It's not real.

 

[TheBigA]

The main reason the data does not match is that both the diary and PPM measure actual listening by the participant. Streams are often left on with nobody listening.

Oh come on. You're the one who always says that diary ratings are inflated because they don't measure actual listening.

PMM doesn't measure listening by the participant, but by the device.

 

[davideduardo]

Oh come on. You're the one who always say that diary ratings are inflated because they don't measure actual listening.

My point with the diary is that interruptions are not "subtracted" since diarykeepers round off listening as is human nature. But streams are often left on when nobody is nearby for hours or days... such as my "entertain the dogs" example. In particular, Alexa users tend to leave the device "making noise" all day long.

PMM doesn't measure listening by the participant, but by the device.

It measures by participant, since the device senses whether it is being carried or not and if not moved it discards the detections.

My point is that streams are not accurate regarding... and usually can't even tell... demographic data, which is the key to ad sales. This is why Nielsen is working to have a total audio service and a total video one as "connections" are nice to the seller of the service, but are useless for most ad budgets.

Or, as we say "a monkey dressed in silk is still a monkey". Streams do not offer adequate data on users.

 

[TheBigA]

My point with the diary is that interruptions are not "subtracted" since diarykeepers round off listening as is human nature. But streams are often left on when nobody is nearby for hours or days... such as my "entertain the dogs" example.

You don't have actual statistics for that, though. Just your anecdotal viewpoint. I for one shut my appliances off when I leave a room.

Streams do not offer adequate data on users.

Somehow or other, all of the big tech companies have figured out how to sell the audience data they have and they're eating radio's lunch. Radio may have better demographics, but Google has the money. I get paid every month from Google and Meta. Radio needs to learn how to sell the growing digital audience or they'll be left like Lotus is now in Seattle, trying to sell a declining AM audience base, and laying off staff.

 

[AmusedObserver]

My point with the diary is that interruptions are not "subtracted" since diarykeepers round off listening as is human nature. But streams are often left on when nobody is nearby for hours or days... such as my "entertain the dogs" example. In particular, Alexa users tend to leave the device "making noise" all day long.

It measures by participant, since the device senses whether it is being carried or not and if not moved it discards the detections.

My point is that streams are not accurate regarding... and usually can't even tell... demographic data, which is the key to ad sales. This is why Nielsen is working to have a total audio service and a total video one as "connections" are nice to the seller of the service, but are useless for most ad budgets.

Or, as we say "a monkey dressed in silk is still a monkey". Streams do not offer adequate data on users.

Hey David, there was a rumor at NAB this year that Nielsen was using paper diaries from former panelists in PPM markets to estimate headphone listening. Sounded crazy to me; you know anything about this? I’d love to knock the rumor down if it’s false.

 

[davideduardo]

Hey David, there was a rumor at NAB this year that Nielsen was using paper diaries from former panelists in PPM markets to estimate headphone listening. Sounded crazy to me; you know anything about this? I’d love to knock the rumor down if it’s false.

I was, as the conventioin program showed, at the NAB in New York 3 weeks ago. There was nothing credible about that specific processes for the development of the headphone algorithm. What was the subject of some talk is the use of panels (not a fast-turnover random sample) to establish earbud listening along. This would include enough data from enough markets to determine usage for all the twenty-plus stratification variables of the ratings survey.

The idea that they might follow up with past metered households / dwelling units makes a lot of sense. These are people who not only "would" be just like today's metered households... they actually were past metered household members. They are also verified participants whose statistics and historical listening are on file.

The main issues involve determining if people, when using earbuds, have similar habits to when they are on speaker-based systems like car audio. That means in what way do they use the same, or, if different, stations on earbuds as they use on speakers. It also means determining the range of times each subset spends listening on earbuds compared to speaker-based systems as well as in what dayparts, days and even holidays they do that.

What they are looking for is the PPM behaviour and an algorithm that says " if measured PPM listening happened in these times to these stations, the unmeasured earbud listening happened in these other specific times and to these stations" with enough measured behaviour from past metered persons to be statistically sound. I think they would have to recalculate the algorithm rather frequently, similarly using recent past metered persons of households.

 

[davideduardo]

You don't have actual statistics for that, though. Just your anecdotal viewpoint. I for one shut my appliances off when I leave a room.

I don't but Nielsen does. It can be obtained very easily by sampling past metered persons and households for "typical behaviour".

Somehow or other, all of the big tech companies have figured out how to sell the audience data they have and they're eating radio's lunch. Radio may have better demographics, but Google has the money.

But much of the money you are talking about goes to visual ads and context sensitive targeting. It does not go to audio streams that are not from radio, to podcasts and other audio products.

Internet ads have replaced what, 90 years ago, radio fought with print for... retail price/item, automotive, real estate and point of purchase.

I get paid every month from Google and Meta. Radio needs to learn how to sell the growing digital audience or they'll be left like Lotus is now in Seattle, trying to sell a declining AM audience base, and laying off staff.

First, you assumption about Lotus is not based on internal decisions and facts. Lotus is a very well financed company with no debt and huge cash reserves. Whatever they are doing is likely part of a plan that none of us have been made party to and which will not, however it goes, endanger a hugely profitable company.

Second, the biggest issue with "selling the digital audience" is that no streamed and ad supported audio service based on music has figured out how to pay performance (artist and label) royalties and make a profit. So if radio companies have not figured it out, they are not alone... nobody else has an answer for free-to-the-listener ad supported audio.

Of course, there is a possibility that we may end up with no free "broadcast" media. It is headed that way in video, but I suspect that before America loses free video or audio, there will be legislation to change the system to allow people without a couple of hundred spare dollars a month for subscription services to have access to entertainment.

 

[Speed Racer 1]

Nielsen has an algorithm that compensates for headphone use of audio. It is based on actual listener behaviour monitoring to create a model for different age groups, ethnicities and the like relative to earbud / headphone listening.

In LA, there are 4,000 meters. The maximum length of usage is 24 months, but the average is less as some dwelling units drop out and others are removed for failure to comply.

That means that out of about 10,000,000 people over age 6 in LA, in the last 15 years of PPM usage, there have been 90,000 metered persons. Do the math: it's really unlikely that you would know anyone with a meter... particularly when those with them are told not to talk about it!

Thank you David for the information. In my opinion this is a ridiculous form of measurement. 4,000 people out of 15 million or so counting the outlining areas. Those numbers are not extrapolated like Neilsen? Like I said, why not just put a chip in each component and get real TSL. I don't see how anyone can make an argument that the PPM is accurately definitely listenership. And I'm new to the board, but I'm im a radio veteran and worked at some of the biggest stations in the US for 2 decades. I just think most of you know more than me and I like to hear your thoughts! I'm grateful for the feedback and appreciate your time.

 

[Speed Racer 1]

After reading this informative exchange, (and I appreciate the input!) I have a belief that the "Nielson"company should be removed and the big radio corps should engage the major audio manufacturers in placing listening-trackability (yes, I just invented that word) receivers in all listening devices. Apps, and everything. I'll tell you WHY we've been using this system for so long ...data. User data. They'd rather have a little data than anything else. Data is all about advertising dollars. So instead of pure TSL from a device chip, we get this. I'd scrap the system and go total TSL from every device. I don't care about the data. I want pure listening data. What would happen? All numbers would be based on 6+ cume. But those numbers would be so big, radio's footprint would re-imerge in the land of audio. We'd be a player again. Oh but we can't do that, we need to sell soccer moms 18-34 Ikea furniture.

BTW, if you don't know your radio stations main demo and how to position it, you shouldn't 't be selling it. How many stations across the US sell advertising yet, don't subscribe. Go total radio device listening and change the world.

How much money does Nielson make? Find that answer and you have unlocked the truth......

 

[TheBigA]

I don't but Nielsen does. It can be obtained very easily by sampling past metered persons and households for "typical behaviour".

It's all speculation.

But much of the money you are talking about goes to visual ads and context sensitive targeting. It does not go to audio streams that are not from radio, to podcasts and other audio products.

Most radio companies do both visual and audio.

Second, the biggest issue with "selling the digital audience" is that no streamed and ad supported audio service based on music has figured out how to pay performance (artist and label) royalties and make a profit.

We don't know for sure because they don't break it out. But digital as a whole is growing, and that's what matters. As for royalties, some companies have negotiated label deals that make it worthwhile. Others combine the streaming with text & audio content they own and manage. They're not streaming for charity. Look at Townsquare.

 

[TheBigA]

After reading this informative exchange, (and I appreciate the input!) I have a belief that the "Nielson"company should be removed and the big radio corps should engage the major audio manufacturers in placing listening-trackability

Do you speak Chinese? All the audio manufacturers are in China. This would have to be negotiated by the federal govt. Not very likely.

Then it would have to get the approval of the advertisers. Remember radio doesn't do Nielsen for itself. It does this because the advertisers insist. They drive the truck, not radio.

 

[bossbill]

Geez, I'm so glad to be in an unrated market. We don't care about the 50 diaries that go out every year in a market of 70,000; the small samples show no consistency. We don't care about PPM because there aren't any in our town. We do care about helping local businesses sell products and services; we work with them, monitor their sales results, tweak the message when necessary.

Thanks for allowing me to barge in.

 

[Kelly A]

Geez, I'm so glad to be in an unrated market. We don't care about the 50 diaries that go out every year in a market of 70,000; the small samples show no consistency. We don't care about PPM because there aren't any in our town. We do care about helping local businesses sell products and services; we work with them, monitor their sales results, tweak the message when necessary.

Thanks for allowing me to barge in.

That's the nice thing about smaller markets. Ratings amount to whether your clients and families listen.

 

[radiofan2023]

After reading this informative exchange, (and I appreciate the input!) I have a belief that the "Nielson"company should be removed and the big radio corps should engage the major audio manufacturers in placing listening-trackability (yes, I just invented that word) receivers in all listening devices. Apps, and everything. I'll tell you WHY we've been using this system for so long ...data. User data. They'd rather have a little data than anything else. Data is all about advertising dollars. So instead of pure TSL from a device chip, we get this. I'd scrap the system and go total TSL from every device. I don't care about the data. I want pure listening data. What would happen? All numbers would be based on 6+ cume. But those numbers would be so big, radio's footprint would re-imerge in the land of audio. We'd be a player again.

Audacy is possibly going to be delisted soon. EMF (one of the only companies buying stations) presumably doesn’t subscribe to ratings. You still wouldn’t be able to sell based on 6+ numbers. This idea is dead in the water.

 

[davideduardo]

It's all speculation.

The exact methodology, yes. The fact that they are doing it is a fact presented to many corporate-level PPM clients.

Most radio companies do both visual and audio.

The visual is supplementary, such as in-car displays of anything from song data to album covers. It is not a product of itself.

We don't know for sure because they don't break it out.

If you have a digital streaming simulcast of your AM or FM, you know the usage data as that is the basis for the streaming digital artist/label royalties. And that is why many smaller groups or local stations geozone the digital content so as not to pay royalties on the stream in areas where they do not get revenueñ

But digital as a whole is growing, and that's what matters.

But the more it grows, the more the royalties / fees cost. So far, nobody, big or small, has been able to achieve profitability with ad supported digital streams.

As for royalties, some companies have negotiated label deals that make it worthwhile.

The only one with significant deals is iHeart, and those deals are awfully (word used advisedly) close to the general rates all stations pay.

Others combine the streaming with text & audio content they own and manage. They're not streaming for charity. Look at Townsquare.

Townsquare does not make its money off streaming. It combines radio with web services customized to the client, not ads on streams. Their setup is almost like an ad agency, where radio is one part of a package that includes a client's web presence and web marketing. More than half their revenue now is non-radio related, but most of all their revenue comes from smaller markets where the clients buy a spectrum of services that usually includes radio.

Townsquare can be seen as in the early phase of transitioning away from radio entirely, offering a broad spectrum of local ad services to clients. Their main competition is the search and business finder sources on the web.

 

[davideduardo]

Thank you David for the information. In my opinion this is a ridiculous form of measurement. 4,000 people out of 15 million or so counting the outlining areas.

It is more than adequate for the needs of the users of ratings, who are ad agencies. When the sample is increased, the precision of the results is not significantly enhanced; they have tried samples that are much larger and there is no real difference except with the occasional very small station... and those stations don't buy ratings and agencies don't buy them anyway.

In any case, that is 4,000 persons every day of the year. In the diary system, there might have been over twice the diaries, but only abut 600 in any given day or week as the "panel" turned over every 7 days. Now, we have thousands of daily panelists every day, enough to give accurate ratings right down to the daypart for one show on one day.

In the PPM markets from 40 to 50 in rank, the panel is 900 to 1000 persons. Still, quite enough for nearly all markets to get MRC certification.

Those numbers are not extrapolated like Neilsen?

We are talking about Nielsen. What do you mean by "extrapolated"? That is not a term I have heard used in ratings services.

Like I said, why not just put a chip in each component and get real TSL.

But then you would not get age, gender, ethnicity, language dominance, income, education data. Just the usage of the device. That is not what advertisers want. Again, the purpose of ratings is to support pricing and ad buying by agencies and big direct accounts.

I don't see how anyone can make an argument that the PPM is accurately definitely listenership.

Ad agencies, via the Media Ratings Council, believe the system is accurate.

Home | MRC

Media Rating Council (MRC) is a not-for-profit industry self-regulatory body, established in 1963 at the request of US Congress, that audits and accredits measurement services.

mediaratingcouncil.org

In fact, it was the MRC and its members who pushed for the PPM back in the 90's. There were agency "big shots" in every meeting on the early PPM development back around 2001-2002 when it started on air testing in Philly. I can say that because I was at those meetings, too, as one of a small group of radio group representatives.

And I'm new to the board, but I'm im a radio veteran and worked at some of the biggest stations in the US for 2 decades. I just think most of you know more than me and I like to hear your thoughts! I'm grateful for the feedback and appreciate your time.

That is why this kind of dialog is interesting and informative.

 

[davideduardo]

That's the nice thing about smaller markets. Ratings amount to whether your clients and families listen.

This is the old "the cash register is all the ratings I need"...

 

[davideduardo]

Geez, I'm so glad to be in an unrated market. We don't care about the 50 diaries that go out every year in a market of 70,000; the small samples show no consistency. We don't care about PPM because there aren't any in our town. We do care about helping local businesses sell products and services; we work with them, monitor their sales results, tweak the message when necessary.

As mentioned before, in local markets advertisers don't have campaigns on all the major TV channels and webs along with multiple radio stations on the buy and supplementary efforts on the web, in devices like mailed shoppers and billboards and other systems like buying facings and end aisle displays in retail locations.

How many different stations or station groups do your clients buy? I'll bet many only buy you and maybe some presence with Google and other web search services. Clients know when ads work, they know who gives them service and helps them with copy, seasonal events and specials. They don't need ratings, they need service.