Mickey Luckoff, President, General Manager of Citadel's San Francisco stations, recently wrote an open letter to Arbitron complaining about PPM. He noted:
One of the main sales advantages Arbitron used in their sales campaign to radio broadcasters was represented to be the ability to measure program elements, tune-in as well as tune-out features. Moreover, when the scant numbers of meter response information is “sliced and diced” the available numbers, trends etc. are of absolutely no value in making any such decision with any degree of accuracy or reliability.
To illustrate Mr. Luckoff’s concern, we looked at one station (we’ll call it WZZZ) in a market that had just switched to PPM. We chose the median station listed in the 6+ ranker. Median means that half the stations had a higher share, and half the stations had a lower share. We then ran a Persons 25-54 analysis of morning drive for the station. (We can’t be more specific than this regarding the station or market without running afoul of Arbitron’s restrictions.)
The first graph shows the number of PPM meters logging the station each minute from 6 am to 9 am. (Click to enlarge.)
The numbers to the left of the graph are the number of meters. The largest number of meters “tuned” to the station was 5. For three consecutive minutes in the 7 am hour, the stations had 5 meters.
The gaps in the graph are those times when the station had no meters registering it. For 58 minutes, or 32% of morning drive, not a single meter logged the station.
The second graph shows the distribution of meters registering “WZZZ”, from zero to five. Note that out of all of morning drive, only five minutes had 4 or 5 meters registering the station.
For fully 70% of morning drive, the station had either 1 or no meters. This was for the median station in the market. Half the market's stations may have had more meters during morning drive, but the other half of the market had fewer.
One of the touted advantages of PPM was the “granularity” of the data. Broadcasters were assured that programming decisions could be made based on minute by minute ratings. Were this station to make changes based on PPM minute by minute, it would be using the actions of just one or two people to determine the outcome.
Now perhaps you can understand why Mr. Luckoff complained that:
Having so few meters stymies the ability to measure any aspect of programming at any particular time. Therefore, the promised advantage to programmers to make decisions based upon "PPM" input is virtually non existent. And for this, our expense for this service increased by 50 percent! This situation will continue to be magnified as Arbitron rolls out the PPM into smaller markets.
Comments