Arbitron about to be impacted by disruptive technology

[Brooklyndon]

http://www.shopkick.com/app.html

This application works in a manner very similar to the ppm. The only difference, rather than highering compliance officers to reward users for carrying the PPM, this app does it automatically. And shopkicks sample size is only going to get bigger...

 

[davideduardo]

http://www.shopkick.com/app.html

This application works in a manner very similar to the ppm. The only difference, rather than highering compliance officers to reward users for carrying the PPM, this app does it automatically. And shopkicks sample size is only going to get bigger...

This does not work like the PPM, since the PPM uses station-side encoding to identify what it is "hearing." There has been exploration of using a station fingerprinting (like MediaBase uses, for example) but then the device would have to record a high enough quality digital audio file and then compare it regularly with a full monitor of the station.

There is also the issue of proportionality... key to MRC accreditation.

 

[Brooklyndon]

This does not work like the PPM, since the PPM uses station-side encoding to identify what it is "hearing."

From the New York Times:
"The app knows someone is in a store by listening for an audio transmitter placed in each participating store; the phone’s microphone picks up the signal, which people cannot hear."
http://www.nytimes.com/2010/08/17/technology/17app.html?scp=1&sq=shopkick&st=cse

To me, given that the company already applies the process in a different format, it seems like a simple engineering task to get an "audio transmitter" put into a radio station, have its transmission included in the broadcast, and then have a device that reads and records that transmission.

This application is a huge threat to the ratings industry, and the rating industry must somehow address that threat.  MRC compliance is very attainable for a well-capitalized start up.  Nielsen or Arbitron should move to buy out Shopkick's investors now.

At the very least, this application, were Arbitron to integrate it into the PPM, could be a very good cost reducer for Arbitron.  But I also see a time, in the very near future, where this application is the standard for producing ratings in non-rural areas.

Lastly, do Arbitron or Nielsen have the capacity to do “store ratings” for malls?  That sort of data is extremely valuable for retailers, because a retailer with a “high rating” will be able to negotiate, based on its traffic generating capabilities, with the mall for better space and cheaper rents.  There is a whole new market for ratings.

 

[davideduardo]

To me, given that the company already applies the process in a different format, it seems like a simple engineering task to get an "audio transmitter" put into a radio station, have its transmission included in the broadcast, and then have a device that reads and records that transmission.

Arbitron's PPM already does this. Other attempts at encoding radio signals have failed as they caused distortion or reduction in the station's coverage, etc (Houston Media Audit test). Arbitron uses patented technology to embed masked coding into each station's broadcast, identifying the station and the time listened. The PPM device detects this code and registers it. Code bursts can occur as often as 12 a minute.

This application is a huge threat to the ratings industry, and the rating industry must somehow address that threat. MRC compliance is very attainable for a well-capitalized start up. Nielsen or Arbitron should move to buy out Shopkick's investors now.

MRC accreditation is enormously difficult to attain. Most of the issues do not have to do with technology but with the recruitment of participants, sample proportionality on all stratification varibles, including age, ethnicity, georgraphy, etc., and the checks and validation of the data.

At the very least, this application, were Arbitron to integrate it into the PPM, could be a very good cost reducer for Arbitron.

Why? They already have a working device that can measure any encoded audio source.

Lastly, do Arbitron or Nielsen have the capacity to do “store ratings” for malls?

Nielson does not do radio in larger markets. They are in a handful of small ones, with generally one time a year surveys.

The PPM can and has done in-store detection, based on encoding of the store or it's chain's in-house audio. It's a measure of point of purchase.

That sort of data is extremely valuable for retailers, because a retailer with a “high rating” will be able to negotiate, based on its traffic generating capabilities, with the mall for better space and cheaper rents. There is a whole new market for ratings.

It is pretty easy to use other methods of traffic counting that don't require each shopper to cary a device around.

And the economics of measuring store traffic are insignficant compared to what stations pay for radio ratings. So, in-store may be an add-on enhancement for the PPM, but not a core business.

 

[K6JHU]

Something has been bothering me. How does PPM work if you are listening thorugh headphones or ear buds?

 

[Brooklyndon]

MRC accreditation is enormously difficult to attain. Most of the issues do not have to do with technology but with the recruitment of participants, sample proportionality on all stratification varibles, including age, ethnicity, georgraphy, etc., and the checks and validation of the data.

It is pretty easy to use other methods of traffic counting that don't require each shopper to cary a device around.

And the economics of measuring store traffic are insignficant compared to what stations pay for radio ratings. So, in-store may be an add-on enhancement for the PPM, but not a core business.

That's the point. People carry their phones everywhere. No more compliance officers. No more expensive focus groups. Volunatary population recruitment, from which it is very easy to take a proportional sample.


The fact that the main point of contnetion with this technology is not whether it is theoretically able to serve the same role as the PPM, but rather whether they can get MCR compliant should be taken as a huge red flag that the ratings agencies have a threat on their flank.

 

[KeithE4]

Something has been bothering me. How does PPM work if you are listening thorugh headphones or ear buds?

And what's to stop someone from being a wise-a** and parking the radio with the PPM device attached on the local Radio Disney or Spanish-language religious station (both usually at or near the bottom of the ratings list) and leaving it there 24/7?

It would be dishonest, but this PPM sounds like it doesn't measure actual listening. It just indicates that a radio with a device attached is turned on and tuned into a given station, regardless of whether anyone is even in the room. How does this help stations and their advertisers?

 

[davideduardo]

And what's to stop someone from being a wise-a** and parking the radio with the PPM device attached on the local Radio Disney or Spanish-language religious station (both usually at or near the bottom of the ratings list) and leaving it there 24/7?

The meter has a motion sensor, and the data processing looks for signs of movement that fit patterns of human behavior; if the meter does not move or move the right way, that day's detections are not tabulated in the survey. So, you can't leave it in front of a radio, and you can't strap it to your dog.

It would be dishonest, but this PPM sounds like it doesn't measure actual listening.

It's not supposed to measure just listening. It is supposed to measure exposure.

It just indicates that a radio with a device attached is turned on and tuned into a given station, regardless of whether anyone is even in the room. How does this help stations and their advertisers?

Considering that the motion detector takes the big absences out of the formula, what we have is vastly superior to the diary, where people wrote down long periods of listening when they actually did lots of other things in that time period.

Again, advertisers pretty much said, "do something like TV has with the meters if you want us to buy radio..." and so we got the PPM.

And in TV, we know folks get up to go to the bathroom or to get snacks during commercials... but all this is factored in to determine the cost of advertising.

 

[davideduardo]

That's the point. People carry their phones everywhere. No more compliance officers. No more expensive focus groups. Volunatary population recruitment, from which it is very easy to take a proportional sample.

First, the app will only work on certain kinds of smart phones. That lets out 90% of the population. Second, there may be certain demos, lifestyles, ethnicities, income levels, etc., that don't have many phones that fit the requirements.

Pollsters don't use self-recruitment because the kind of "active participant" that signs themselves up is not necessarily representative of the universe in any group. Almost all valid research is recruited very scientifically with many levels of stratification and that means self-recruits will not produce a valid sample.

Focus groups are not used to do ratings. In radio, focus groups are done to find out why one morning show does better than another... touchy-feely stuff, not quanitative stuff.

In case you did not get the big picture, the system you are so sold on is principally intended to encourage people to buy, shop and visit certain retail locations. It's an electronic loyalty rewards card.

The fact that the main point of contnetion with this technology is not whether it is theoretically able to serve the same role as the PPM, but rather whether they can get MCR compliant should be taken as a huge red flag that the ratings agencies have a threat on their flank.

The MRC does not have anything a service has to "comply" with. The MRC grants, or doesn't, accreditation. Since the MRC is made up of media and advertising experts and its goals are to insure that ratings are an accurate buying tool, there can be no single approved system. The Arbitron diary has been accredited in all but one US market for the last several decades... The PPM system still has issues regarding sample proporionality that prevent all but three PPM markets from being accredited.

The reason the PPM is not yet on telephones is that the current system works uniformly, as the meter is of a single type with a single specification. Going on a cell phone means that the differences in microphones, sensitivity, directionality, battery life, etc., would make each phone variety different... thus throwing a horrible variable into the process. For ratings to work via phones, it is likely that the phones would have to be a limited range of devices, all tuned to the same specs, and provided by the ratings company for as long as the participant is part of the survey. The alternative is an unpredictable opportunity for bias and distortion and even fraud.

 

[davideduardo]

Something has been bothering me. How does PPM work if you are listening thorugh headphones or ear buds?

The PPM has a little headphone pass through.

 

[Tom Wells]

Something has been bothering me. How does PPM work if you are listening thorugh headphones or ear buds?

The PPM has a little headphone pass through.

But then it would be measuring listening, not exposure. Or does the mic stay open while your listening to "X", and recognize that you're
being exposed to "Y"? What if you're carrying a Boom-box tuned to "X", turned up loud....will the PPM still register exposure to "Y"?
Will it register two at once? What is the detect/reject ratio in db?
Was this issue even addressed?
What if the user is listening to a recorded show (with or without PPM encoding), does it still register, and does it get counted?
Again, what about the extraneous, unwanted exposure....should PPM carriers always listen with the radio up loud, to ensure
that the device does not recognize the exposure to an unwanted signal?

 

[davideduardo]

But then it would be measuring listening, not exposure.

Listening is exposure.

Advertisers and ad agencies want to know who hears... who is exposed... to each ad. They don't give a darn if the person exposed to the ad picked the station, although there is an advertising and radio industry committee working with Arbitron to create an audience commitment index that will also show how engaged listeners are with stations they hear... a new metric for radio.

Or does the mic stay open while your listening to "X", and recognize that you're being exposed to "Y"? What if you're carrying a Boom-box tuned to "X", turned up loud....will the PPM still register exposure to "Y"?

The PPM detects at roughly the threshold of intelligible hearing. If you can hear the audio, so can the meter and if it can hear the station, it can extract the PPM tags which can occur as many as 12 times a minute if there is audio to mask the 4.5" data tag.

Will it register two at once?

I wouldn't know, and have never asked. The logical answer is a question: "how often does a person put up with two audio sources at the same time?" Since few people do, and if it happens it is for such brief intervals as to render the outcome statistically irrelevant that I really don't care.

What if the user is listening to a recorded show (with or without PPM encoding), does it still register, and does it get counted?

Yes, if the show is recorded off the air or off a stream that is encoded, TiVo like crediting will occur within a window of time... and the credit goes to the time of the recording, not listening as the PPM tag identifies the station/stream and the time broadcast.

Again, what about the extraneous, unwanted exposure....should PPM carriers always listen with the radio up loud, to ensure that the device does not recognize the exposure to an unwanted signal?

That's not, as I said, the objective. The PPM is indended to determine the exposure to advertising and radio stations and streams as people go about their normal routine... and if you change their habits, the measurement is not representative any more. Plus, with the average panelist staying on the panel about 11 to 12 months, nobody is going to go around blasting the radio for a year.