Arbitron admits it doesn't know how to measure satellite radio

[zumahans]

http://www.mediaweek.com/mw/news/recent_display.jsp?vnu_content_id=1003156833

From Media Week:

“We recognize the implementation date for reporting satellite in local markets has moved previously and for this reason we hesitate to provide another target survey date until we have this desired additional data,” said Brad Feldhaus, vp of product management and client services for Arbitron, which hopes to begin reporting satellite radio channels sometime in 2007.

(go read the article and click on the advertising)

 

[OldGringo]

http://www.mediaweek.com/mw/news/recent_display.jsp?vnu_content_id=1003156833

From Media Week:

“We recognize the implementation date for reporting satellite in local markets has moved previously and for this reason we hesitate to provide another target survey date until we have this desired additional data,” said Brad Feldhaus, vp of product management and client services for Arbitron, which hopes to begin reporting satellite radio channels sometime in 2007.

(go read the article and click on the advertising)

Excuse me, but direct from the Arbitron Advisory Council discussions we know that:

1. Arbitron has been mesuring satellite for XM for several years and they know how to measure satellite listening, or XM would not be paying millions a year for it.

2. The issues are based on wording and the physical layout of columns in the diary and instructions to the idarykeeper. All suc changes have to be approved by the MRC for implementation, and such approvals take a while.

3. The Advisory Council also submits its recommendations which are "considered" and form part of the finally implemented decision.

Since satellite represents less than 1% of all listening, and the ad market is not the same as that for local radio, this is hardly a burning issue.

 

[clichemoth]

The results of this should be interesting when fully implemented. I suspect alternate delivery systems haven't shown up before only because listeners don't know that they can report them.

Yes, there won't be enough listeners to each individual channel out of the hundreds on Sirius and XM and the countless thousands on the Internet to amount to much. But that isn't really the point. The point is to show how many people are listening to these other forms of radio in total. It's probably more than most people think, especially since they'll be counting streams of terrestrial stations separately from actual OTA listening.

The people who didn't grow up with OTA radio as their primary music choice will constitute the majority of 25-34 within a few years. We've adopted new technology, we buy things online, being wired is an integral part of our existence just like radio was to those in the era of the classic top-40 stations. Local advertisers still need to reach us, but how will they do it and will streaming audio be a vehicle for them to do so?

 

[OldGringo]

The results of this should be interesting when fully implemented. I suspect alternate delivery systems haven't shown up before only because listeners don't know that they can report them.

I think, after looking at many markets and too many diaries that listeners do wriet down XM and Sirius, as both market as "satellite RADIO" so the meaning is pretty clear.

If you do the math, you also see that the current share potential of satellite is well under a 1 share... more in the 0.4 to 0.6 range, depending on whether you think satellite users totally abandon terrestrial or whether they share in different proportions.

Yes, there won't be enough listeners to each individual channel out of the hundreds on Sirius and XM and the countless thousands on the Internet to amount to much. But that isn't really the point. The point is to show how many people are listening to these other forms of radio in total. It's probably more than most people think, especially since they'll be counting streams of terrestrial stations separately from actual OTA listening.

Eventually, radio will have to make sure streams are counted with the terrestrial station. The impediment right now is that very few streams of importance are 100% simulcasts, due to AFTRA related commercial removal. Radio has to work with agencies and AFTRA so that the onerous restrictions of the union are changed to something more reasonable.

The people who didn't grow up with OTA radio as their primary music choice will constitute the majority of 25-34 within a few years. We've adopted new technology, we buy things online, being wired is an integral part of our existence just like radio was to those in the era of the classic top-40 stations. Local advertisers still need to reach us, but how will they do it and will streaming audio be a vehicle for them to do so?

The real issue is fragmentation. How can you get a stream or two that have any impact on local audiences? I don 't think either satellite or streaming is the killer app against radio. It's probably WiMax, which can be localized and is "radio-like" in all aspects.

 

[SuperRadioFan]

Will the PPM (?) devices measure online streaming and/or WiMax listening data and if so, how?

 

[OldGringo]

Will the PPM (?) devices measure online streaming and/or WiMax listening data and if so, how?

Any encoded source, even storecasting, will be measured. Online streams will have separate coding.

 

[zumahans]

Interesting. How will onstream be encoded separately? If a person buys a CD and broadcasts it on the web, how does the PPM know it wasn't a CD that was compressed and played on a MP3 player.

 

[ricoontheradio]

Interesting. How will onstream be encoded separately? If a person buys a CD and broadcasts it on the web, how does the PPM know it wasn't a CD that was compressed and played on a MP3 player.

Because that CD is not encoded with the proper digital encoding that is talking to the PPM... I assume?

 

[OldGringo]

Interesting. How will onstream be encoded separately? If a person buys a CD and broadcasts it on the web, how does the PPM know it wasn't a CD that was compressed and played on a MP3 player.

Each audio "source" including TV, is encoded "at origin" (in the case of radio, in the audio chain) with a unique code. So the WZZZ on air signal has one code, and the streamed audio is encoded with a different code. Same for each TV and cable channel, each internet-only station that wants to participate, and each satellite channel.

If a streaming station is encoded, the code is from the stream, not the indvidual audio components to the stream. CDs are not encoded, nor are commercials (although commercials could be encoded in additon to the station encoding to get ratings for each commercial).

The PPM detects the codes, and saves them in memory with a time stamp (each source sends the code many times a minute) and this data is downloaded from the recharger / dock each night to Arbitron.

 

[OldGringo]

Interesting. How will onstream be encoded separately? If a person buys a CD and broadcasts it on the web, how does the PPM know it wasn't a CD that was compressed and played on a MP3 player.

Because that CD is not encoded with the proper digital encoding that is talking to the PPM... I assume?

Yep. If one wants to check song airplay, they use MediaBase or BDS which use previously created fingerprints of each song to identify in real time the spins for each tune on monitored stations.

Signals, not content, are encoded for PPM.

 

[zumahans]

Very interesting. So if a kid rips a CD, sends the MP3 to his friend (illegally), and then streams it using any one of the dozens of homebrew streaming programs, what shows up on the PPM?

And I take it the PPM has a mic that picks up the encoded audio somehow? Or are radio listeners going to have to indicate something?

 

[OldGringo]

Very interesting. So if a kid rips a CD, sends the MP3 to his friend (illegally), and then streams it using any one of the dozens of homebrew streaming programs, what shows up on the PPM?

And I take it the PPM has a mic that picks up the encoded audio somehow? Or are radio listeners going to have to indicate something?

The PPM has a microphone which is calibrated in sensitivity to the same level as the average human ear. If a person hears the sounds, the PPM does, and extracts the code every time it "hears" it and reccords it in Flash memory with a time stamp.

Again, the station encodes itself... no matter what is on the air. Commercials, jock chatter, traffic reports, music, sports... the encoding is added to the audio just ahead of the transmitter in the audio chain. Every minute of broadcast has multiple code microbursts that can be detected by the PPM.

No CD is encoded. Nothing is encoded in the production room at the station. Encoding is done right ahead of the transmitter. Networks don't encode... the affiliates do. An internet station would encode right ahead of the T1 that delivers the stream to the backbone. Satellite encodes every channel separately in the Sirius or XM TOC, not in the studios.

The CD you refer to does not register at all unless the internet station's stream is encoded. And int his case, the internet station is encoded, not the CD or MP3 itself.

 

[zumahans]

What's to stop a radio station for amplifying this subaural tone and playing it at, say, a mall, where PPMs will pick it up?

 

[OldGringo]

What's to stop a radio station for amplifying this subaural tone and playing it at, say, a mall, where PPMs will pick it up?

The encoder is a "black box" inserted into the audio line, so getting the tone would be somewhat difficult.

The main deterrant is that a market like LA will have around 2.40o meters. The chances of one being at a specific mall are low, and the time that it would detect a station would also be limited to the time the person was in the mall halls, as stores have thier own background music. All you would get would be maybe 10 or 15 minutes of listening; P1 listening per staiton is often above 10 hours weekly.

This was brought up at the Arbitron PPM fly-in two weeks ago in Columbia. the premise was that a football stadium would play "encoded" music for the pre-game period. With 50,000 in attendance, in a market of 5 million, the chances are that there might be 5 to 12 meters in the stadium and each would detect for a brief period. This would not be enogh quarter hours to change the outcome of a survey by even 0.1 share points, and the cost for such an insignificant gain would be prohibitive.

There is less probablility of this happening than an occasional station getting a few diaries and filling them in.

 

[zumahans]

---->The encoder is a "black box" inserted into the audio line, so getting the tone would be somewhat difficult.

Give me a motive, 10 minutes, a computer, an oscilloscope and a station silent for a few minutes at 4 a.m. on a Mongday (attn: Akbar), and I can name that tone.

Then, let's say, I have a friend who works at the dub house. He starts laying my tones in on all sorts of other commercials and dubs.

Another friend lays it over the LA Times commercial that plays at thousands of movie houses every day.

Then, I have a friend at the music label who lays it onto a song that is distributed for free via MP3s.

Whammo. I am in the Los Angeles MRS.

 

[OldGringo]

---->The encoder is a "black box" inserted into the audio line, so getting the tone would be somewhat difficult.

Give me a motive, 10 minutes, a computer, an oscilloscope and a station silent for a few minutes at 4 a.m. on a Mongday (attn: Akbar), and I can name that tone.

The tone is not inserted in silence. So there must be an input for it to "encode." It would be very difficult to find and isolate the microbursts.

Then, let's say, I have a friend who works at the dub house. He starts laying my tones in on all sorts of other commercials and dubs.

Since it is pretty near impossible to isolate the microburst, the amount of work required and the chances of success do not merit it. Further, it is a clear violation of the Arbitron contract, so if this is found out, the station will be delisted... something that could cost a major market station millions. And the PPM will only be in major markets.

Another friend lays it over the LA Times commercial that plays at thousands of movie houses every day.

Ratings work in quarter hour increments. To get credit for any specific quarter hour, the meter has to continuously detect for 5 full minutes. So a one-minute commercial that is encoded will not give credit to anyone.

Then, I have a friend at the music label who lays it onto a song that is distributed for free via MP3s.

The 3 or 4 minute song does not meet the 5 minute minimum. No credit will be given.

Whammo. I am in the Los Angeles MRS.

There is no such thing as "an MRS" or the "Los Angeles MRS." Minimum Reporting standards is a cut off criteria that determines which stations are "in the book" for the ratings.

While you cited the rule, you did not understand that MRS means Minimum Reporting Standard and requires a certain minimum num ber of diary mentions and a certain ammount of listening in them to be included in ANY market report.

And the MRS rule you cited appleis to diaries. There are different minimum levels of listening that apply to the PPM.

At the end of the day, the real issue is the trouble and expense and the risk vs. gain.

1. Trouble and expense: paying to have the station played in a location with many people. I consider extrracting the microburst to be very unlikely as it only comes out of the box if there is audio. It does not encode silence.

2. Risk. If you were able to isolate the burst, you violate the contract and will be delisted if discoverd. Some disgruntled emplyee eventually will turn you in.

3. Gain. In LA, the average listener spends 20 hours a week listening to radio. There will be about 2400 meters. That is 48,000 hours a week or about 200,000 quarter hours. In the case of a stadium, with one dial for every 5000 persons, you might have 8 dials that are worn by spectators who are from inside the market. And lets say you put the station on or encode (were you to be able to do it) the stadium audio, you would get maybe 8 total hours of pergame detection. That is 1/100th of 0.1 share points. Considering what you would have to pay, there is no measurable gain to be had.

It's in violation of a contract, it is ineffective, and it gets no measurable gain.

 

[zumahans]

You do it again.

I said Los Angeles MRS - Minimum Reporting Standards - and I meant Minimum Reporting Standards.

As in, my unrated radio station is now in the book. Through some audio trickery I am in the book.

I didn't say the market, I said it meets the standards.

David, why do you keep doing this? Haven't you been humiliated enough in the San Diego board?

 

[ricoontheradio]

What would be interesting (and probably only realistically possible in small markets than those like LA) is stations making GREAT friends with, for example, local car dealerships. A nice handshake that every car on their lots will have your station on. Then those test-driving, walking the lot, etc. are being recorded.

This will obviously also increase the game of "listen at work" by having businesses keeping your station on. PPM holders walk into those businesses can add a quarter hour here, quarter hour there. Again, insignificant in a market like LA, but a market like Fresno it might be a little more effective to maybe make a difference here and there.

 

[zumahans]

What about places like hair salons, doctors offices, restaurants, etc.

All kinds of public places could skew the rating.

If I hear Movin at the car wash, I plug my ears, but the PPM records it.

 

[ricoontheradio]

What about places like hair salons, doctors offices, restaurants, etc.

All kinds of public places could skew the rating.

If I hear Movin at the car wash, I plug my ears, but the PPM records it.

Right, but again, I don't like markets like LA will be effected much by that. Smaller markets (When they finally get PPM in 3-5 years), maybe. Would be interesting to see.