January book is out

[four4rock]

Weighting is only used to make the sample totally proportional to the actual population. When people under 30 are "weighted up" is when the sample got too few of them and, to be proportional, each one that they got is given more weight.

There is no weighting for listening times or even for listening at all. Weighting is done to achieve balance for individual age groups, gender, geography, education, race and income.

Your points are well made. As to the PPM, its greatest effect would be to reduce the time spent listening on average by about 40% due to much greater precision than the rounded-up entries diary keepers tend to make. That would reduce the average audience size, and for buyers seeking cost per point or CPM goals, rates would drop dramatically.

For Western New York and in a typical (or on average) how much weighting is actually done? For example are there sample segments that are effectively doubled or tripled... or is it far more fine 'tweaks' in the tenths or hundredths?

Also, after umpteen cycles, I'd assume that disproportionate response rates can be generally predicted... so why not increase the solicitation for those historically non-responsive groups?

 

[CTListener]

For Western New York and in a typical (or on average) how much weighting is actually done? For example are there sample segments that are effectively doubled or tripled... or is it far more fine 'tweaks' in the tenths or hundredths?

Also, after umpteen cycles, I'd assume that disproportionate response rates can be generally predicted... so why not increase the solicitation for those historically non-responsive groups?

I'm guessing it's because that would be pouring money down the drain on something the advertisers and agencies haven't expressed a need for. They're convinced that the numbers produced through weighting are valid enough for their purposes, not "voodoo math," as Mr. Shula calls it.

 

[davideduardo]

For Western New York and in a typical (or on average) how much weighting is actually done? For example are there sample segments that are effectively doubled or tripled... or is it far more fine 'tweaks' in the tenths or hundredths?

Generally, weighting is in the single digits and is sort of "sanding the rough edges". But in some cases, we may find 15% to 20%, and during the pandemic, of course, there was a greater chance of the need to weight.

In PPM markets where we get "authorized" weekly reports we see stations that are quite stable from month to month (with no more than 10% wobbles) having individual weeks that are considerably more varied.

Also, after umpteen cycles, I'd assume that disproportionate response rates can be generally predicted... so why not increase the solicitation for those historically non-responsive groups?

The problem in diary markets is that the sample is new every week. So, as returned diaries from two or three weeks back come in, adjustments in the placement for upcoming weeks are made. And, like a pendulum, the sample is always swinging back and fort due to that. A participant is contacted and recruited for a future week based on time to process, time to mail the diary(ies), and margin of time for postal delays.

The cost of each participant is very high. While the incentive is small in the diary, the cost to recruit, follow up with multiple phone calls and tabulation is great. Excess sample will not solve the issue but will vastly increase subscriber cost. Nielsen already know what percentage of people in each subset will actually fill in a diary and return it, and they know which groups will accept a recruit effort so their work is calculated with all those facts... even seasonal variations... in mind. In effect, they already "over-recruit" by sending out more diaries to groups that have a lower return rate.

So a sample excess or lack is not known until the diaries for that week are all returned and tabulated, at least 20 days. Then the recruiting for about 2 to three weeks in the future is adjusted, meaning that it takes 4 to 6 weeks to adjust. Every week brings new adjustments. And, finally, weighting takes place.

They are trying to get proportional samples for each age cell, gender, race/ethnicity/language preference (for Hispanics), income, education and county of residence. There is also an effort within counties to get balance base on sub-areas that can be classified, although that is not a "guaranteed and weighted" goal (it's called "Geozones" by Nielsen).

And, by the way, it may seem that users are complacent or totally satisfied by the ratings, but we are constantly making comments to Nielsen about all kinds of issues (I'm talking about "bitching" here). Many broadcast groups have in-house experts... some former Arbitron / Nielsen high-level staffers... who are looking at the data constantly. Remember, subscribers get a huge array of possible reports and there are add-on services that get even more deep data.

 

[davideduardo]

I'm guessing it's because that would be pouring money down the drain on something the advertisers and agencies haven't expressed a need for. They're convinced that the numbers produced through weighting are valid enough for their purposes, not "voodoo math," as Mr. Shula calls it.

The weighting process in the diary markets is ongoing, as is constant sample adjustment during each "book". This was part of the original Arbitron radio sample design by Jim Seiler back in the mid-60's. It was further refined in the very early 70's when Arbitron standardized survey periods nationally and it was necessary to try to bring the sample into proportionality by each survey period's end back when measurement was not continuous.

 

[four4rock]

Generally, weighting is in the single digits and is sort of "sanding the rough edges". But in some cases, we may find 15% to 20%, and during the pandemic, of course, there was a greater chance of the need to weight.

In PPM markets where we get "authorized" weekly reports we see stations that are quite stable from month to month (with no more than 10% wobbles) having individual weeks that are considerably more varied.

The problem in diary markets is that the sample is new every week. So, as returned diaries from two or three weeks back come in, adjustments in the placement for upcoming weeks are made. And, like a pendulum, the sample is always swinging back and fort due to that. A participant is contacted and recruited for a future week based on time to process, time to mail the diary(ies), and margin of time for postal delays.

The cost of each participant is very high. While the incentive is small in the diary, the cost to recruit, follow up with multiple phone calls and tabulation is great. Excess sample will not solve the issue but will vastly increase subscriber cost. Nielsen already know what percentage of people in each subset will actually fill in a diary and return it, and they know which groups will accept a recruit effort so their work is calculated with all those facts... even seasonal variations... in mind. In effect, they already "over-recruit" by sending out more diaries to groups that have a lower return rate.

So a sample excess or lack is not known until the diaries for that week are all returned and tabulated, at least 20 days. Then the recruiting for about 2 to three weeks in the future is adjusted, meaning that it takes 4 to 6 weeks to adjust. Every week brings new adjustments. And, finally, weighting takes place.

They are trying to get proportional samples for each age cell, gender, race/ethnicity/language preference (for Hispanics), income, education and county of residence. There is also an effort within counties to get balance base on sub-areas that can be classified, although that is not a "guaranteed and weighted" goal (it's called "Geozones" by Nielsen).

And, by the way, it may seem that users are complacent or totally satisfied by the ratings, but we are constantly making comments to Nielsen about all kinds of issues (I'm talking about "bitching" here). Many broadcast groups have in-house experts... some former Arbitron / Nielsen high-level staffers... who are looking at the data constantly. Remember, subscribers get a huge array of possible reports and there are add-on services that get even more deep data.

Crazy stuff. Sounds like there's enormous opportunity for improvement on may levels. It is what it is, I guess. Must be one heck of a disclaimer in the purchase/subscription contract!

 

[Buddy Shula]

OK. So people are stupid and can't remember what station to write in the diary. That's not Nielsen's fault. It's also possible that most Radio content is not worth remembering...

Thank it appears we must remind them to write it in the diary they receive

 

[Buddy Shula]

You make a valid point about Radio listening. Many people have no idea what they're listening to. They barely pay attention. The content is not compelling and is just background noise.

Buddy claims he would prefer PPM. He would not like the results at all. He would then rail against that system. The PPM would not accurately report WECK due to technical difficulties. His 5 share guarantee still hasn't happened so it must be Nielsen's fault...

Over 5 shares have happened often in our demos

 

[davideduardo]

Crazy stuff. Sounds like there's enormous opportunity for improvement on may levels. It is what it is, I guess. Must be one heck of a disclaimer in the purchase/subscription contract!

The simple disclaimer is "audience estimates".

Any improvement costs money. That cost is passed on to subscribers. If you owned or managed a station, how'd you like your ratings contract to go up by 50% next year?

 

[tbolt909]

Over 5 shares have happened often in our demos

Good of you to clarify that you meant "In Demo and not 12+". That's not what you've said in the past. It really shouldn't matter. The demos aren't going to change whether you have a 3 or 5 share 12+. All that matters is if your clients and listeners are happy(and if you're business is thriving)...

 

[davideduardo]

You make a valid point about Radio listening. Many people have no idea what they're listening to. They barely pay attention. The content is not compelling and is just background noise.

Most people write the dial position in the diary, followed by station names. Most people have a very definite idea of what they are listening to, and can name their three or four favorite stations.

Buddy claims he would prefer PPM. He would not like the results at all.

He would not like the 40% reduction in actual people, on average, listening to his station and to radio. That is what happened in 2009-2010 when most PPM markets rolled out.

He would then rail against that system. The PPM would not accurately report WECK due to technical difficulties.

What "technical difficulties"? The PPM is a very simple system of an embedded audio data stream. If a person can hear the station, the meter can hear the "code".

His 5 share guarantee still hasn't happened so it must be Nielsen's fault...

He gets that all the time in 55+. And he seems to be able to sell that demo locally.

 

[tbolt909]

He would not like the 40% reduction in actual people, on average, listening to his station and to radio.

What "technical difficulties"? The PPM is a very simple system of an embedded audio data stream. If a person can hear the station, the meter can hear the "code".

You obviously missed the sarcasm. Buddy has railed against "Weighting" and the Nielsen methodology. He would not like PPM in Buffalo. You have tried to explain it with no success. He doesn't trust Nielsen...

 

[Rusty Bridges]

Thank it appears we must remind them to write it in the diary they receive

Any type of reference to ratings or "dear diary," "write it down" or "make a note" on-air reminders creates red flags at Nielsen. In the past, stations airing those types of 'reminders' have been sanctioned or listed 'below the line' for such antics.

 

[Buddy Shula]

Any type of reference to ratings or "dear diary," "write it down" or "make a note" on-air reminders creates red flags at Nielsen. In the past, stations airing those types of 'reminders' have been sanctioned or listed 'below the line' for such antics.

These are "Antics" - they are the truth. The station is not owned by Nielson. Either is the radio industry. If I want to tell people to "write it down" to get their vote, I am going to.

I would take that to court in a second if Nielson every played around with me. Why should they need to approve the marketing of our station. They want people to send back diaries? We will tell them to send them back.

 

[four4rock]

These are "Antics" - they are the truth. The station is not owned by Nielson. Either is the radio industry. If I want to tell people to "write it down" to get their vote, I am going to.

I would take that to court in a second if Nielson every played around with me. Why should they need to approve the marketing of our station. They want people to send back diaries? We will tell them to send them back.

Man, I'm on board with you 100% here. In the absence of a direct contractual stipulation, it seems that such a "prohibition" is a gross overreach.

 

[Rusty Bridges]

Man, I'm on board with you 100% here. In the absence of a direct contractual stipulation, it seems that such a "prohibition" is a gross overreach.

And there you have it.

 

[TheBigA]

Hey Buddy Shula! I thought of you when I saw this. Byron Allen has sued Nielsen for fraud, among other things:

https://tvnewscheck.com/regulation/article/allen-media-files-multi-billion-dollar-lawsuit-against-nielsen/

Byron Allen’s media companies Weather Group, Entertainment Studios Networks (ESN) and CF Entertainment filed suit against Nielsen seeking billions of dollars in damages for “fraudulent misrepresentation and fraud by concealment.”

The lawsuit asserts that there has been tremendous fragmentation in the industry and proliferation of channels, rendering the Nielsen panel ratings system antiquated, highly unreliable and fundamentally flawed.

Also, the suit claims, Nielsen told Allen’s companies that its panel system could reliably rate ESN even though the networks had limited distribution at the time. The lawsuit alleges that Nielsen knew, but failed to disclose, that the panel system was unreliable across all networks, and that the system was totally unreliable for networks like the ESN.

The lawsuit further alleges that Nielsen concealed these facts and thereby caused Entertainment Studios and the Weather Group to pay millions in fees. The lawsuit identifies several other networks that similarly received fundamentally unreliable services from Nielsen.

Byron owns The Weather Channel and a few other things. He lawsuit sounds a lot like what you say here.

 

[davideduardo]

Any type of reference to ratings or "dear diary," "write it down" or "make a note" on-air reminders creates red flags at Nielsen. In the past, stations airing those types of 'reminders' have been sanctioned or listed 'below the line' for such antics.

Or de-listed.

 

[davideduardo]

These are "Antics" - they are the truth. The station is not owned by Nielson. Either is the radio industry. If I want to tell people to "write it down" to get their vote, I am going to.

I would take that to court in a second if Nielson every played around with me. Why should they need to approve the marketing of our station. They want people to send back diaries? We will tell them to send them back.

It's in your contract to follow "ratings distortion and bias" rules.

 

[davideduardo]

Man, I'm on board with you 100% here. In the absence of a direct contractual stipulation, it seems that such a "prohibition" is a gross overreach.

It's in the contract that ratings rules will be followed.