PPMs delayed in major markets

[rho01]

Arbitron, beset by complaints that the new people meter methodology is riddled with inaccuracies, has postponed expansion of the PPMs in New York by nine months, LA and Chicago by six months and San Francisco, San Jose and Dallas by three months. Arbitron has not said how this will impact Atlanta's rollout, set for fall of 2008, though it's reasonable to assume it will be delayed here as well.

People meters have been in place in Philly and Houston, revealing huge increases in cume, big drops in many cases in TSL, especially for ethnic and religious stations. Some radio groups have complained that Arbitron has been undersampling minorities and 18 to 34 year olds, the same problem they've had with the old diary system. Apparently, asking a 21 year old to carry a pager is no better than asking them to fill out a paper diary. And obviously, trying to REACH a 21 year old nowadays is tough given that more and more of them have ditched the landline and live off cel phone alone. And PPMs are showing evidence that minorities don't listen as much to certain stations as they say they do.

Rock and pop stations seem to have benefited in Houston and Philly.

Obviously, by measuring actual listening, PPMs provide more accountability than diaries and advertisers naturally prefer that. Some buyers accuse the media companies of stalling because the stations know that under the new system, ratings aren't as high and thus, they may be charging more than they should for their product.

Here's a good Ad Week story about the issue:

http://www.adweek.com/aw/magazine/article_display.jsp?vnu_content_id=1003676883

 

[FCR]

Arbitron's press release says Atlanta's roll out won't be affected, but time will tell.

http://ftp.media.radcity.net/ZMST/ARBPPM11-26.pdf

 

[SalesWeasel]

Why would anyone under 25 listen to the radio anyway? PPM is the best thing to happen to ratings; it levels the playing field and gives advertisers a real account of listening.

 

[RoddyFreeman]

Arbitron does need to get the sampling correct, especially in the 18-34 cell.

As far as urban goes, however, I have heard from radio people who have done diary reviews that urban stations have a high incidence of people writing in the call letters for the first quarter hour of listening, and then drawing a line all the way down through the day (and night). In other words, they feel the diary has been overstating listening to urban stations and that the PPM methodology is not the problem with the reduced urban station ratings.

 

[davideduardo]

As far as urban goes, however, I have heard from radio people who have done diary reviews that urban stations have a high incidence of people writing in the call letters for the first quarter hour of listening, and then drawing a line all the way down through the day (and night). In other words, they feel the diary has been overstating listening to urban stations and that the PPM methodology is not the problem with the reduced urban station ratings.

Actually, all kinds and ages of listeners do the line down a daypart or day thing. It is particularly common in AC stations during work hours, or for high profile morning shows. This is by no means a Black or ethnic thing.

Additionally, there is no "quarter hour" place in a diary. There are a bunch of blank lines for each daypart, 6-10, 10-3, etc. The listener fills in the start and stop time, the station (usually the dial position) and whether it is AM or FM and where the listening took place. So someone who listened to an AC from 9 AM to 5 P will write in 9 AM dnd draw a line and then write in 5 PM in afternoons and the station info.

In the PPM, the meter shows toilet and coffee and lunch breaks, and that fact is not format specific. I've looked at tens and tens of thousands of diaries over the last 37 years in Beltsville, Columbia and Laurel... and I don't see rounding and line drawing as being limited to any one group of listeners.

 

[deadman]

so when in 2008 does Atlanta start PPM?

 

[davideduardo]

People meters have been in place in Philly and Houston, revealing huge increases in cume, big drops in many cases in TSL, especially for ethnic and religious stations.

That is a bit broad. There are meters in Houston, Philly, New York, Chicago, Riverside, LA and San Francisco. Only Houston and Philly were "currency" and NY was to be currency in 5 weeks.

The increases in cume came from short interval, occasional or even accidental listening. In the ppm, 92% of quarter hours comes from less than half the cumers. So half the cume is useless, except for driving down the average TSL because the "bad half" only listens for a few minutes a week.

In Houston, the last Diary survey showed 19.5% of shares going to Spanish language stations. In the October PPM, there were 19.7 Spanish language shares. SO the "loss" of ethnic shares is exaggerated in many cases.

The issue is that the sample is not proportional on every stratification variable... and that means the results are not accurate.

Some radio groups have complained that Arbitron has been undersampling minorities and 18 to 34 year olds, the same problem they've had with the old diary system.

That is a big exaggeration. Hispanics in the diary sample have, for the most part, been truly proportional to their incidence in the population. 18-34 requires, like Blacks and Hispanics, DST procedures to get proportionality in the sample, and that means Arbitron just works harder for those diarykeepers. But in most cases, the sample is quite proportional and minimally weighted. In the PPM, there are cells 40% below proportionality... and a panel is supposed to be fully proportional on every variable... that is why weekly results would be as accurate as a month or a quarter or a year. But if the panel is not proportional, the results in the more granular analyses are flawed.

Apparently, asking a 21 year old to carry a pager is no better than asking them to fill out a paper diary. And obviously, trying to REACH a 21 year old nowadays is tough given that more and more of them have ditched the landline and live off cel phone alone. And PPMs are showing evidence that minorities don't listen as much to certain stations as they say they do.

The panel is based on households and not individuals. 21 year olds are not recruited. Family units are. Every member is expected to participate, and if one does not, the whole family is dropped from the panel. The inability to keep a balanced, proportional and participatory panel is what is plaguing Arbitron, who has never done panel based research before for radio ratings. They obviously did not know how to do it as well as they thought, or they would not have blinked today.

Rock and pop stations seem to have benefited in Houston and Philly.

And AC stations and classic hits (oldies) several other formats.

Obviously, by measuring actual listening, PPMs provide more accountability than diaries and advertisers naturally prefer that. Some buyers accuse the media companies of stalling because the stations know that under the new system, ratings aren't as high and thus, they may be charging more than they should for their product.

Listening has not changed... advertisers are getting just as many listeners per spot as ever. When the PPM is ready for prime time, we can analyze listening levels in this new system... which will probably be more precise and reflect more accurately what has been the truth all along but that the diary could not capture. For the moment, though, the diary reigns.

 

[davideduardo]

so when in 2008 does Atlanta start PPM?

Pre-currency was to start in October, 2008. It is unlikely that will happen given today's news.

 

[Salsa Shark]

So much for those stations waiting for PPM to "save" them. My money is on PPM not making it to Atlanta until 2009 now...

 

[DashRiprock]

Is the growing Web based audience considered when setting rates these days? It’s a lot easier to track downloads/streams to specific I.P. addresses than diaries or PPM systems.

 

[bgwilde1]

David, you are spot-on!

I know at least part of the grumbling stations are doing....

Stations LOVE to do physical diary audits if they think something funny is going on. With the PPM, it's going to be much harder for stations to look at goofy handwriting, locate misplaced slogans, or dispute the validity of an ink stain.

PPM may be more accurate than stations would like to hope...

 

[seabass]

Arbitron, beset by complaints that the new people meter methodology is riddled with inaccuracies, has postponed expansion of the PPMs in New York by nine months, LA and Chicago by six months and San Francisco, San Jose and Dallas by three months. Arbitron has not said how this will impact Atlanta's rollout, set for fall of 2008, though it's reasonable to assume it will be delayed here as well.

That's because the problem with PPM is that it tells the truth, which is more than radio executives can handle these days.

They've got enough headaches without now revealing to media buyers that they've strung them along for years with 30% less audience than they've been selling, and in the case of urban stations, even more inflated audience estimates.

It certainly makes for a good charge of fraud, if it weren't for the fact that radio stations are vouchsafed by Arbitron's lousy methodology with the paper diaries. But still, radio keeps shooting itself in the head and compounding mistakes by blaming Arbitron.

As a group, I don't think there have been businesspeople with this little vision and adaptive skill since money was invented.

 

[RoddyFreeman]

Actually, all kinds and ages of listeners do the line down a daypart or day thing. It is particularly common in AC stations during work hours, or for high profile morning shows. This is by no means a Black or ethnic thing.

So Mr. Eduardo, to what do you attribute the huge--and I do mean huge--drop in New York for all of the urban stations with the PPM methodology? African-American listeners are always oversampled, and from what I've read, there was not a sampling problem with the Black audience in New York. Yet AC stations, which you say also have a high incidence of getting a line drawn from the start of listening through the day, were not hurt.

The people who told me about this are too senior and knowledgeable to be wrong. Unless you can quantify your statements, I'm going to continue to believe those people.

 

[davideduardo]

So Mr. Eduardo, to what do you attribute the huge--and I do mean huge--drop in New York for all of the urban stations with the PPM methodology?

Bad sample, bad carry rate, bad proportionality. That is why the NY PPM could not be accredited by the MRC.

African-American listeners are always oversampled,

No, that is totally untrue. All groups, by age, sex, geography, language dominance (Hispanic), and ethnicity are very close to proportional in the diary survey, and any minor differences are adjusted up and down by weighting.

and from what I've read, there was not a sampling problem with the Black audience in New York.

There are huge sample issues in 18-34, and since both Blacks and Hispanics are younger on the average, the very low proportionality of 18-34 (only about 60% of target) affects those groups the most.

Yet AC stations, which you say also have a high incidence of getting a line drawn from the start of listening through the day, were not hurt.

That is because the cume increase from "hearing" as opposed to "listening" balanced with the loss of TSL on P1 and P2 listeners to reach fairly equivalent shares.

The people who told me about this are too senior and knowledgeable to be wrong. Unless you can quantify your statements, I'm going to continue to believe those people.

Since you have posted a significant error of fact about "oversampling Blacks" I suggest you may be talking to the wrong people.

The fact that Arbitron pulled NY and is holding back LA, Riverside, Chicago, Dallas and San Francisco / San Jose indicates there is something very, very wrong. The company lost over 20% of its market value in the last 30 hours, and I would assume that management will not be getting bonuses this year and next, at least... plus most top management is likely seeing their stock options under water. A decision of that magnitude indicates a problem of similar magnitude as evidenced by the inability to get NY and Philly accredited.

 

[davideduardo]

That's because the problem with PPM is that it tells the truth, which is more than radio executives can handle these days.

If the PPM were showing "the truth" it would have been accredited by the MRC in NY and Philadelphia. It is not accredited. The sample is not right, the cells are not proportional, and the recruitment and maintenence of the panel is in trouble. That is hardly "the truth."

Were radio afraid of the PPM, then we, as an industry, would not have supported the project and financed it by means of much increast costs. Radio pays nearly all the costs for ratings... why would we pay for something we do not want?

They've got enough headaches without now revealing to media buyers that they've strung them along for years with 30% less audience than they've been selling, and in the case of urban stations, even more inflated audience estimates.

The diary was the only viable method of measurement for decades. Arbitron started work on the PPM in the mid-90's and has done lots of testing on the technical system. The issues, though, are not technical.

We all know that people do not write down short incidents in a diary. And they round off listening periods, often to the hour and half hour. The PPM measures by the minute... and of course real listening will be lower and precise. We have known that for more than half a decade, since the first tests of the PPM in the US. Stations and advertisers get the same amount of impact, as listening has not changed, just the way it is measured. Stations do not suddenly have less listeners... the measurement is simply more precise.

TV went through this some time ago. The initial Nielsen efforts were not the best, and there were lawsuits and more. Finally, they got it to a point where subscribers are in agreement.

To show this is not a minority or ethnic issue, the last diary book in Houston had 19.5 Spanish langauge shares and the PPM in November has 19.7 shares. There is about a point difference between Black shares in the diary and the PPM, so no big change there. There were no "inflated shares" there.

[/quote]It certainly makes for a good charge of fraud, if it weren't for the fact that radio stations are vouchsafed by Arbitron's lousy methodology with the paper diaries.[/quote]

The development of the motorcar did not make horses "lousy" but did make people less likely to buy a horse. The diary is not "lousy" since up until recently there was no other effective measurement method for radio to use. There is certainly no fraud, as the diary methodology has been and still is MRC accredited, something that the PPM can not lay claim to.

But still, radio keeps shooting itself in the head and compounding mistakes by blaming Arbitron.

Nobody is blaming Arbitron for listening levels. We are all, collectively, asking for a well done accredited report with true proportionality in all stratification variables and as close to 100% DDI compliance as possible.

As a group, I don't think there have been businesspeople with this little vision and adaptive skill since money was invented.

There is no evidence to this claim. Arbitron promised, but has still not fulfilled. The MRC, which basically is there to guarantee accurate methodology and measurement for advertisers, is not liking the PPM... but they still approve the diary based surveys in nearly 300 markets.

 

[middlega]

They've got enough headaches without now revealing to media buyers that they've strung them along for years with 30% less audience than they've been selling, and in the case of urban stations, even more inflated audience estimates.

Although I haven't seen it addressed very much here, speaking as one of those buyers, this creates an interesting situation of another sort: we have to reassess what levels are necessary for a campaign to be effective.

I mean, if a buyer has settled in around (just for an example) 500 GRP's based on the old methodology and that was effective but now we discover that those 500 points were probably just 300 points all along then the buyers have to rethink what their GRP levels should be in order to remain effective.

In short, the numbers expected vs delivered really don't seem to be that incredibly horrifying IF the campaigns were effective at whatever the "real" level turns out to be. It's just that we buyers have to recognize that we probably don't really need the levels we thought we did (not when they're expressed based on different currency).

 

[LewisDRoberts]

OK, So now Arbitron admis flaws with the diary methods and now admits flaws with PPM. Where does that really leave them? If I were running an ad agency, I would not use Arbitron to set price per point. It is all flawed.

 

[davideduardo]

They've got enough headaches without now revealing to media buyers that they've strung them along for years with 30% less audience than they've been selling, and in the case of urban stations, even more inflated audience estimates.

Although I haven't seen it addressed very much here, speaking as one of those buyers, this creates an interesting situation of another sort: we have to reassess what levels are necessary for a campaign to be effective.

I mean, if a buyer has settled in around (just for an example) 500 GRP's based on the old methodology and that was effective but now we discover that those 500 points were probably just 300 points all along then the buyers have to rethink what their GRP levels should be in order to remain effective.

In short, the numbers expected vs delivered really don't seem to be that incredibly horrifying IF the campaigns were effective at whatever the "real" level turns out to be. It's just that we buyers have to recognize that we probably don't really need the levels we thought we did (not when they're expressed based on different currency).

Thank you for a very candid and interesting assesment of the issues on the buying side. This has to be about the best PPM post I have read.

 

[davideduardo]

OK, So now Arbitron admis flaws with the diary methods and now admits flaws with PPM. Where does that really leave them? If I were running an ad agency, I would not use Arbitron to set price per point. It is all flawed.

I do not see where Arbitron admits "flaws" in the diary system; that methodology is MRC accredited and audited and woks as well as possible within the limits of the "technolgy" and the sample size.

Buyers of advertising, such as agencies and large accounts, have been asking for electronic measurement because it should be, when done correctly, more accurate and much more timely and easy to use.

The issues with the PPM have to do with getting the service running with a proportional panel on every stratification variable and keeping panelists over time and insuring that they participate (carry) daily. The NY test data showed this was not yet perfected... thus the lack of accreditation.

 

[InTIMadate]

Told you the PPM was junk research. The trades today are full of stories about how Arbitron is on the ropes for selling radio their smoke and mirrors. Beware the board posters here that are in Arbitron's pocket. They will fill your head and mouth with false tales and false facts about PPM. This is not about how "radio should be supporting PPM." It's how Arbitron should come clean and improve their sample before selling it at a premium as fact.

I welcome the coming congressional investigation where Arbitron will have to expose their dealing with the Media Rating Council. There’s something wrong with their methodology. There’s precedent for such a request. Congress quizzed Nielsen about its Local People Meter. It's time to expose their dark secrets to the bright light and find out how ugly this baby really is.

How did our radio leaders let Chuck DuCoty end up being our voice with Arbitron as the leader of the Arbitron Advisory Council? He's the guy that took Q101 Chicago from billing $23 Mil to $16 Mil while sinking it to the bottom of the ratings. He's lucky to have a job in Cedar Rapids. We need real leadership watching these oily salesmen not a spotlight hog.

Arbitron is back-pedaling faster than a Democratic front-runner on the military surge. It's time to give them a swift kick and teach them the meaning of the word "Respect" in order to protect radio from their heavy-handed, monopoly robber-baron efforts.

Can you guess why so many programming professionals are losing their jobs this quarter? Corporate radio must cut jobs to pay for the 65% increase in Arbitron fees for the PPM in a year that is showing a loss in advertising revenue. Sad..... so sad.