ATTORNEY GENERAL CUOMO'S REACTION TO ARBITRON'S RECENT ACTION

[Mark Giardina]

A friend of mine sent this press release to me on Monday dealing with Andrew Cuomo and his "crusade" against Arbitron. I thought many of you might find this interesting reading. Feel free to comment.



"Arbitron's decision to release its unreliable and unaccredited radio

ratings system is an affront to racial and ethnic minorities in New York and

around the country. The Attorney General's Office cautions all advertisers

and broadcasters against using these prematurely released ratings as we

believe they are flawed and will be the subject of ongoing litigation.

Arbitron's unwillingness to defend the validity of their system on its

merits proves it places its own economic incentives over the interests of

minority broadcasting. As a monopolist, Arbitron owes consumers an

explanation for its decision to force feed the PPM system to broadcasters.

Their irresponsible decision threatens the existence of diversity in radio

and muzzles the voices and viewpoints of millions of Americans. Obviously,

the Attorney General's Office will continue to seek justice in this case."

 

[SirRoxalot]

ArbitRage

Gee, you'd think that Wall Street was a better target for our State Crusader this week. How much is Arbitron going to take out of the state treasury compared to the thieves who created the latest financial crisis?

Arbitron is far from perfect, but the PPM is more perfect than the diary system. Arbitron is a "monopoly" simply because no other methodology is as good as theirs.

 

[scooterodell]

Would someone please explain to me how the use of PPM is racist?

 

[Bob1370]

Racism has been alleged because the sampling method seems, when compared with years of past diary results, to squeeze down the AQH and cume of urban format and Latino format stations, and boost the numbers of conservative white-oriented talkers and other formats unlikely to appeal to black and Latino listeners.

I don't know if that's true, and certainly no one can prove deliberate bias based on what we now know about the ratings game. One thing we do know, is that there is a very wide disparity between the numbers shown for various formats in the diary surveys, and the numbers that have come up to date in the PPM surveys. This was true in markets that have had PPM for a while, and is now true in the newly released NYC PPM report as well. Certain formats, including classic hits, CHR and conservative talk, seem to do much better in PPM than in diary surveys take through this spring...and in the case of conservative talk, do quite a bit better than you'd expect in the markets surveyed which are anything but conservative in the way they vote. Other formats, like urban, urban AC and Latino, do significantly worse than the demos of the market would lead you to expect. Of all the formats, only non-partisan talk, sports and straight all-news seem to show the same share (within the margin of survey error) by either method. Everyone else's numbers seem to deviate sharply from the patterns of the recent past.

This tells you that something's wrong with at least one of the methods, if not both.

Is it the diary method that's off, and PPM giving a better snapshot of listener preference? Or is PPM off, especially given that it yields results that seem counterintuitive given the demographics of the markets where it's in use? And which method gets you the best take not only on what people hear, but what programs make an impression on your target demo and may help you as an advertiser?

One clue to why there is a difference is that the diary method measures basically voluntary listening, the stations that the survey respondent chose to spend time with and recall later when filling in the diary. That's useful for advertisers because they get to know who's choosing to spend time with the shows they buy, and where they might be making an impression that can lead to a sale. Conversely the PPM catches not just voluntary listening but also coincidental listening, which can include involuntary and even unwanted and resented accidental listening to stations you might never choose and may actively dislike but were exposed to when you got into a cab, or walked into a store, or had to sit in a doctor's office. This is both PPM's strength and weakness. It catches whatever the survey respondent is exposed to. But it doesn't distinguish between desired listening by choice, which makes a positive impression for the advertiser supporting it, and involuntary listening, which may be resented and cause either a negative impression of the advertiser or no impression at all, resulting in a total waste of the advertiser's dollar. I'll admit if I carried a PPM and were exposed in a store to a program and host I deeply dislike, I'd form a negative impression of the advertiser and make a note not to patronize him--not to mention walking out of that store playing the show, without making a purchase. PPM won't catch any of that, and it's crucial for the advertiser to be able to infer. it will just mark me as a listener, even to a program I might detest and which makes me feel animosity toward whoever advertises on it.

One of the virtues of a diary was that listeners who formed such negative impressions, if they chose to record that listening at all, were able to note them down, and stations looking at the diaries were able to make note of that qualitative feedback. That's knowledge that a smart advertiser and a smart station manager will want, and if they get enough feedback like that, programming lineups have even been known to change. They won't get that knowledge with PPM unless it's accompanied by a parallel diary to collect that qualitative info.

The other thing we need to know, is how the sample for each survey method was picked, and how it's corrected for response rate fluctuation. Are we getting a good snapshot not only of listening habits, but of the people who are using radio in the market being surveyed?

I don't think we're getting solid information on any of these questions. Until we do, we need at very least to tell Arbitron to run both methods in parallel, and test each one for accuracy in sampling selection, adequacy of sample size, and their ability to deliver the qualitative info that really makes the raw numbers meaningful to both advertisers and stations. There is NO ratings method in use at the moment, that really gets the job done in the way advertisers and broadcasters need it done. We need to either push Arbitron to get it right, or put together a consortium of advertisers and broadcasters to lead us to a new and better method.

 

[aaronread]

There is NO ratings method in use at the moment, that really gets the job done in the way advertisers and broadcasters need it done. We need to either push Arbitron to get it right, or put together a consortium of advertisers and broadcasters to lead us to a new and better method.

And while you're at it, why not have Arbitron perfect cold fusion while bringing peace to the Israeli-Palestinian conflict? You want the impossible, man: you cannot possibly measure the IMPRESSION a person gets from listening to a SPECIFIC ad. Human psychology just doesn't work that way - it's like Heisenberg's Uncertainty Principle: trying to increase the resolution to the point of just one commercial in every possible setting decreases the objectivity of the results. They become too subjective to be quantifiable across a wide enough sample audience to be statistically significant. And anyone who presents findings that claim they have achieved this Holy Grail is not exactly lying, but you cannot take what they say as gospel (pardon all the religious puns) since the statistics will be so pliable that you can make the data say whatever you want.

To pair this back to your comment...how exactly would Arbitron EVER find out that you heard a commercial on a radio station playing in a store where you were buying something? Much less find out that you hated the commercial so much you'll never shop there again. Hmmm, the only way they could determine that is if you self-reported it. And to be fair, you'd have to write it down...the closer in time to the actual occurrence, the better. You'd probably have to carry around this small black book and a pen to write it all down...just like a...err...hmmm...yeah, a diary. Oops. The wildly subjective and inaccurate self-reporting system is exactly what the PPM is supposed to get AWAY from.

Anyways, what this is really about is not racism...it's about money. The complainers even say as much in the NY Daily News:

Frank Flores, general manager of WSKQ, says unofficial PPM ratings released over the last year show La Mega "has dropped 50%-60% in its average quarter-hour share. I don't know if we can survive that."

That's right Frank, it's all about the benjamins. I have little sympathy for a station that learned how to game the diary system to their advantage (probably by programming to attract high loyalty from a smaller audience) that suddenly finds out their audience isn't nearly as big as they deluded everyone - even themselves - into thinking. And it's not like we all haven't had several years' warning that the PPM often finds that diary reports are wildly inaccurate.

I don't think the PPM's results are exposing flaws in the methodology nearly as much as they are shattering many comfortable illusions that radio stations have had about listening trends and styles.

 

[SirRoxalot]

Real Listening

Anybody who isn't aware that promotion is at least as important as programming for Aribitron diaries isn't paying attention. There are a significant number of people who write down what they "should" like instead of what they DO like. It ain't hip to like Barry Manilow, but he sells millions of units. Neither methodology really addresses "active" listening vs. "passive" listening.

The PPM more accurately records what people actually hear. There is a significant push in minority communities to get people to write down ethnic stations whether that's what they actually listen to or not. In every day life, people are more "American" than "____-American". I don't find this to be a bad thing.

There are problems with Arbitron sample sizes, and audience composition. In fact, Arbitron has likely oversampled minority and ethnic populations, and seriously undersampled the transient (no land-line) population under 30. This is true of whatever method they use.

Andrew Cuomo is only getting involved in this national issue because:

1. He wants his name in the paper so he has visibility on a national scale (see the "Spitzer Model").

2. There's political gain to be made among minority businesses and audiences.

Cuomo's angling for bigger things. This is one way to get his name out there and get talked about.

 

[Bob1370]

"And while you're at it, why not have Arbitron perfect cold fusion while bringing peace to the Israeli-Palestinian conflict? You want the impossible, man: you cannot possibly measure the IMPRESSION a person gets from listening to a SPECIFIC ad."

Not what I'm talking about, or why I have a problem with the ratings we're getting by BOTH diary and PPM methods. I'm saying that each of them has strengths, and potentially fatal weaknesses. The weakness of the diary method was that it depended on memory, and may have missed exposure to programming that was either forgotten or deeply disliked by the survey participant (something a programmer would care about, but an advertiser might not, since a listener who didn't get his message may not be a listener he's worried about). The strength of the diary method was that it gave stations and advertisers a good sense of what station, which program, made enough of an impression on a listener to get him to make a record of it. It also gave survey participants a chance to give some qualitative feedback on the shows to which they were exposed, and that was useful supplemental information that put a person's listening choices in context and gave a hint to how the listener responded to the programming and, by inference, to the advertiser. I can tell you first hand that stations and programmers took that information seriously, made a point to go down to Maryland every quarter to parse the diary comments, and used that info in making programming decisions and figuring out how to sell programs and personalities based on how people responded to them.

You're not going to get that from PPM UNLESS you supplement it with parallel diary surveys, or other comparable research. PPM is designed to give you an instant snapshot of whatever audio the survey participant is exposed to, whether it's a welcome bit of entertainment or an annoying interlude that creates a negative impression and does an advertiser no good. On the level of telling you who happened to make a chance encounter with a radio show at a given moment, maybe it's more accurate if the survey sample was designed properly and is large enough that there's no major error in sampling validity. It's strong in compiling total exposure assuming the sampling is good--but gives you no context for that listening.

You need to find a way to do two things...bring back the kind of qualitative, supplemental information that diary comments have given you to put the raw numbers in context, and get the best of what the overall exposure of a station PPM measures can give you to find out where you're being played and who's encountering you, intentionally or accidentally.

Currently we have problems we need to resolve. First, the two survey methods don't at all correlate in terms of how they measure most formats, and PPM is delivering numbers that don't always seem to make sense in the context of market demographics or what we know about how programming formats appeal to certain demographics and turn others off. Either years of ratings, accompanying qualitative surveys and voluminous focus group research in every large market were all essentially wrong in what they told us, or we're getting numbers from the PPM system that don't correspond to the impressions our programming efforts are really making with the audience. We need to know which is the case. Either we were screwed up for the past 50 years of Arbitron measurement in what we thought we knew about the audience, or we're screwed up now.

Cuomo's lawsuit threat, which is undoubtedly politically motivated, is almost beside the point. The one good thing it has done, is fuel a discussion of the potential weaknesses of all the audience measurement methods we now have. The solution is probably a mix of both measurement methods, combining coincidental station impressions from PPM measurement with a diary or other detailed survey of what kinds of programs listeners actually seek out and respond to. Then perhaps we can either mix the numbers together to get an arithmetic average, or weight them in some manner, to get an actual measure of what people are really tuning in and paying attention to. Isn't that what advertisers and programmers really need to know?

 

[scooterodell]

While I can see the value of the qualitative data that the diary method affords, it still is limited. Most advertising money is spent by agencies, and most agencies aren't interested in how loyal a listener is or whether or not they like the station they're hearing. The want the raw numbers: how much of the population am I reaching with the spot, and how much is it gonna cost me. For that kind of info, PPM makes the most sense.

As for the purported racism in the use of PPM, the results of PPM (as presented in this thread) make sense to me. Urban stations have some of the highest listener loyalty there is; a subjective diary entry from a highly loyal listener, regardless of format or ethnicity, is likely to skew somewhat inaccurately in the station's favor. Conservative talk tends to get a lot of play in businesses and workplaces, where people who don't necessarily agree or line up with the expected demographics are going to hear it. And yes, the industry has for a long time programmed for the diary. You can make statistics say what you want them to... it is likely that we're getting a clearer picture of how things are with the PPM. I really doubt that Arbitron would have intentionally chosen a racially imbalanced sample to use the PPM. What would they have to gain from that?

 

[aaronread]

Cuomo's lawsuit threat, which is undoubtedly politically motivated, is almost beside the point. The one good thing it has done, is fuel a discussion of the potential weaknesses of all the audience measurement methods we now have.

Horsefeathers! We've debated the PPM ad nauseum by this point. Cuomo's making us all go round and round some more and come back to the exact same point: while the PPM ain't perfect, it's a HELL of a lot more accurate than diaries ever could hope to be...and inevitably that's going to provide a rude awakening for some stations. Oh well, folks...live by the ratings, die by the ratings. People still underwrite WEOS even though we can't really point to ratings; we use other means to demonstrate audience popularity and loyalty. Works pretty well for us.

 

[scooterodell]

Oh well, folks...live by the ratings, die by the ratings. People still underwrite WEOS even though we can't really point to ratings; we use other means to demonstrate audience popularity and loyalty. Works pretty well for us.

And that's not just limited to non-comms. WYSL rarely gets much credit from Arbitron, but Mr. Savage seems to be making it just fine.

 

[Bob1370]

"Cuomo's making us all go round and round some more and come back to the exact same point: while the PPM ain't perfect, it's a HELL of a lot more accurate than diaries ever could hope to be...and inevitably that's going to provide a rude awakening for some stations."

I'm not at all sure that's true.

New York City is a glaring example of how PPM is delivering numbers that are counter-intuitive and defy any expectation that the market's demographics and voting behavior would lead you to expect. Maybe it makes sense for a straight-ahead CHR to be #1...we've seen that in NYC for almost 50 years now, it's a format that by its nature tends to bridge a lot of demographic divides. But in a community where African Americans and Latinos dominate demographically and the community politically is deep blue, would a red-state-style station like WABC, representing views most New Yorkers don't share, be expected to post a sudden 40% surge in AQH numbers the moment the methodology changes, and surge into the top 5? Doesn't make sense... and much as age 40-60 folks like me enjoy classic hits stations like WCBS-FM, you wouldn't expect it to score with the 25-34 crowd, it's not playing their music but PPM says it does. Odd...

 

[scooterodell]

I still think that comes down to the fundamental difference between a diary and PPM. PPM records what audio you are exposed to, not what you choose to listen to.

Because it's friendly to the 40+ crowd (lots of managers and small business owners there) and not necessarily a turn-off to their customers (in all age ranges), a station like WCBS is likely to be a popular at-work choice, or the choice of the guy who controls the radio in the small business. The same may well be true for WABC... meaning that you have a lot more people hearing WCBS and WABC as compared to the number of people who are actively tuning the station in. The diaries would likely have just recorded those people who were actively tuning in... and that's not a true picture of who is hearing the station. Whether or not those people are paying any attention or not is irrelevant to the PPM. It's what they're exposed to that matters. And I don't see anything so far that is beyond the realm of plausibility. Should it be looked at more closely? Probably. Should it be a subject of interest to the AG? I don't think so.

 

[SirRoxalot]

Getting to the Point

The real question that I have is this:

Should the Attorney General of NY be getting involved in a matter that concerns radio stations, their advertisers, and a ratings service. I don't see the public anywhere in this matter, so why is government getting involved in the first place?

If Arbitron methodology is that flawed, neither the radio stations nor the advertisers will pay for it. There are other ratings services, and other methodologies for determining ratings. Arbitron has already been stymied by lack of acceptance in several markets. Why not let the market determine winners and losers here?

Instead, hundreds of thousands of your tax dollars are going to AG staffers who are combing through boxes and boxes of documents, completing studies on methodologies that don't directly affect the public.

This is all an attempt by Andrew Cuomo to get his name in front of a national audience. The merits of this investigation are dubious, at best.

 

[Mike Sheridan]

I think Mr Cuomo should be worrying about other things...

 

[exradio]

would a red-state-style station like WABC, representing views most New Yorkers don't share, be expected to post a sudden 40% surge in AQH numbers the moment the methodology changes, and surge into the top 5? Doesn't make sense... and much as age 40-60 folks like me enjoy classic hits stations like WCBS-FM, you wouldn't expect it to score with the 25-34 crowd, it's not playing their music but PPM says it does. Odd...

Those are exactly the kind of things I would expect. If you're 25 and you walk into a store or into the next office and WCBS-FM is playing in the background, and it's not your music, and it's on a station not marketed to you, and you don't know what it is (even regular listeners barely know what station they're listening to), it's unlikely you would write that down in a diary, but with the PPM you're recorded as listening.

If instead that radio is playing WABC, and you're not familiar with it, (consider that a majority of people under 40 in America are not familiar with any AM station), you're not going to write that down in a diary either. With the NYC market having as many stations as it does, WABC's new and improved share is still much smaller than the share of Republican voters.

This all leads to the question of how much an advertiser wants to pay for a station where a listener is paying little or no attention. Formats that show higher AQH with PPM may find advertisers demanding a lower CPM.

 

[SirRoxalot]

Hear Hear

I agree with scooter and exradio in their assessment of the difference between diary numbers and PPM numbers. It's much easier to manipulate diarists. PPM records what they actually hear. Whether they "hear" a station or "listen" to a station is a fundamental point of discussion.

Smart buyers look at both cume and quarter-hours, and tailor buys to those numbers. If you're doing a saturation campaign with lots of ads, you look at cume. If you're targeting specific listeners with fewer, well-placed ads, you look at quarter-hours. The programming content - whether it's background or foreground, ad placement - where it falls in a spot cluster and how big that cluster is, and other factors will determine how effective the ad is.

The bottom line is that advertisers who do a lot of radio have found it to be an efficient, cost-effective way to advertise as long as their spots were memorable, and they targeted the right audience.

 

[JustPastBuffalo]

...in a community where African Americans and Latinos dominate demographically and the community politically is deep blue, would a red-state-style station like WABC, representing views most New Yorkers don't share, be expected to post a sudden 40% surge in AQH numbers the moment the methodology changes, and surge into the top 5? Doesn't make sense...

Be careful about ascribing Red State-Blue State characteristics when considering New York City ratings, WABC in particular as well as the Classic Hits and Classic Rock stations licensed to New York City.

The NYC Metro is as unique as it is large, incorporating counties in New Jersey (Middlesex-Somerset-Union) and Long Island (Nassau-Suffolk), which also stand as separate MSAs in their own right.

These areas have ethnic and demographic variables, different, though not entirely, from Manhattan, Yonkers, Bronx, Brooklyn or Staten Island.

When analyzing ratings for New York City and WABC in particular, it might be helpful to consider Shares vs. Rating. WABC, according to Radio & Records PPM ratings review, garners a 4.8 Share, listeners 12+1.

Now, only for the sake of discussion, let's assign a 2.0 Rating to WABC's 4.8 12+ Share.

The 4.8 Share is based on persons aged 12+ listening to radio (1NB: While R&R attributes these PPM ratings to listeners 12+, I am under the impression that PPM tracks listeners 6+, but I'll defer to R&R's ascription), whereas the 2 Rating is based on the entire 12+ population (ibid).

WABC may be attracting a substantial number of listeners that may have Red State characteristics within a larger population base that may exhibit Blue State characteristics.

In this regard, WABC does a superb job of knowing who their audience is and giving them what they want to hear. Using modern CHR programming concepts, WABC plays the hits.

In the context of audience Share (actual listeners), 4.8 is a significant number. In the context of Rating (population), 2.0 is a smaller number, but a very loyal "choir" and no doubt a 2 rating is worth a lot of money in Market #1.

In a political sense, WABC is serving its base, which is considerable in the context of Share, but not as large in the context of Rating. Clearly, WABC listeners are loyal and dedicated to the station and its personalities and the station makes money (or so we presume, Citadel's share price not withstanding.)

Personally, I'm not fond of the politics of WABC or its personalities, but this isn't a discussion of politics as much as it's a discussion of ratings, research and programming.

-JPB

 

[aaronread]

But in a community where African Americans and Latinos dominate demographically and the community politically is deep blue, would a red-state-style station like WABC, representing views most New Yorkers don't share, be expected to post a sudden 40% surge in AQH numbers the moment the methodology changes, and surge into the top 5? Doesn't make sense... and much as age 40-60 folks like me enjoy classic hits stations like WCBS-FM, you wouldn't expect it to score with the 25-34 crowd, it's not playing their music but PPM says it does. Odd...

Not odd at all. Like I've been saying, the diaries are a hideously flawed product if the goal is to measure actual listenership to a given station, be it AQH or Cume. There's a reason why courts rarely rely on eyewitness testimony anymore; it's too easily manipulated - the memory is too pliable. It's quite believable that if you program to target a niche audience and that audience loves you, that they'll "remember" listening to you "all day" in a diary, when a PPM might realize that "all day" is really only an hour or two here and there.

Also, there is a tremendously flawed assumption being trotted out time and again in this case: just because there are more (given minority) people living in (given locale) and I program to (given minority) then it makes no sense that PPM would show that I have fewer listeners in (given minority) audience. The two items are weakly related at best...so what if you've got ten-thousand more Latinos in NYC this year than last (to pull a number out of thin air). If 75% of them are under 20, and most of them don't drive, why would you assume they're listening to the radio? They're probably listening to iPods - not radio.

The point here is that for all the "it just doesn't make sense" arguments surrounding the extreme trend shifts that PPM reveals, there are just as many perfectly-good rationales to explain that the PPM is right and diaries were wrong as there are vice versa. Given the ridiculous inherent inaccuracies to the diaries, and the inherent accuracies of the PPM, I'd be more inclined to trust the PPM.

The only way I would be inclined to assume the PPM is "wrong" is if the encoding is somehow being boogered on the station in question. A technical glitch is a perfectly understandable cause, albeit if you're in the number 1 market and you're not paying top dollar to make sure your PPM gear works right 100% of the time, then you don't deserve to run that station! If ever there was a "don't eff around with this one" aspect of engineering, PPM would be it!

Of course, a technical glitch also would completely disqualify any reason Cuomo could give for wading into this...

 

[chrish]

Coumo is another braindead socialist who thinks the government is the answer to everything.......Arbitron is a private business, if the stations do not like the methodology the are free not to buy it.
Another firm will come along and give them what they want.

Government gave us Barney and Herbies Fannie Mae and Freddy Mac adventure
along with Chris Dodd by threatening financial instituations with loss of their charter
is they did not make sub prime loans to people who had no way of paying them back.

Coumo is just the beginning.......I'm afraid Americans have been dumbed down so
far [EDIT]

Americas founding fathers must be rolling in their graves just watching what
dispicable politicans judges have found ways to to pervert the Constituions
meaning.


[EDIT-off topic politics]

 

[Element9]

Rare is the occasion that I stand behind the mods. It is, however, their board and "rules is rules." Can we leave the wingnut rhetoric outside the door, please? Cuomo's not a socialist, he's an AG tilting, in this case, at the wrong windmills. Barney (as inane as the program may be IMHO) isn't corrupting the minds of little children. (Their parents are, by setting them in front of the TV and not reading books to them.) And as far as the Constitution and our founding fathers (who by the way, were quite a group of rogues in their own right) are concerned, there are other places to better discuss The Federalist Papers.

Cuomo is playing this "Arbitron PPM alledged discrimination" situation all wrong. I suspect he's been getting a good deal of "homespun information" from the owners (the poor dears) of radio stations in NYC that cater to Hispanics and African Americans. He may need to better distill that information by reviewing the expert opinions of the esteemed professionals, wannabees, poseurs and ex-cons(cripts) who post on this board. Heh! A little humor on a Sunday morning.

If you want to read posts by a very knowledgeable professional, I suggest you seek out those of Dave Eduardo (aka, Dave Gleason) who frequently writes on the Programming, NYC and LA boards. Not always do I agree with him, but the man has an extensive background in broadcasting and research, and he's usually right.

As to Arbitron? It's a big company with a lot of money and a firm grip (one might argue a monopoly) on the balls of the radio business. Looking at the stock market and the potential cutbacks that may result from the downturn in the broadcasting business, I wonder if any groups and clusters will be re-structuring or flat-out canceling their contracts with Arbitron. Nah. Easier to cut personnel.

-9-