Phantom Cume

Hey, Merry Christmas! Sorry for the long delay since the last blog entry. I just did an analysis - should have done this years ago - that shows the fall music testing season extends almost precisely from September 15 to December 15. At that time, some things definitely get put on the back burner. And, I regret, this blog has been one of them. But I'll try to be a lot more regular, at least until the spring music testing season explodes on March 15.

This week, Arbitron (the US ratings company) held its annual "Consultant Fly-In". And they released some "interesting" information.

First, some background. When PPM was introduced, the first and absolutely huge difference everybody noticed was the greatly elevated cume level for many stations. At the same time, TSL dropped like a submarine with screen doors.

Arbitron's John Snyder addressed this. Using the Houston PPM data, he showed that most of the extra cume now being seen actually comes from people who listen to a station less than one hour per week. In his words, "..they had exposure."

The radio trades are reporting this, but they aren't connecting the dots. Let me help.

1. The diary did a fine job of measuring a station's cume. As the Ball State University study showed, diaries lose about 5% of listening. Put another way, you lose mostly short listens, from people who aren't seeking you out, but just happen to wander into your audio field.

2. Maybe that lost listening is enough to mean something to the sales department. But it means very little to those of us in programming.

3. TSL is an average. Include more very short listens, and it drops. Nobody likes short TSL. The odds of missing a commercial go up, and advertisers go away. But this should be easy to explain, since we are, in truth, only adding an extra, formerly unmeasured, part of our audience. We are adding those who "had exposure."

4. But sadly, radio is probably in trouble with this one. We don't have many sophisticated tools for dealing with audience TSL distribution. The ratings book doesn't have a separate cume value for "greater than 1 hour per week" listeners. And all of our existing reach and frequency curves, originally developed by Westinghouse in the mid-1070's , are now useless.

5. In fact, the diary has been kind to radio. It loses a little listening, but it is from people who are engaging in behavior that is worthless to advertisers. Now those advertisers will beat us up over our new low TSL values. Radio? No loyalty!

6. Worse, we are going to get what we asked for, a better look at reality. A lot of people don't really use or care much for radio. Those people were, I think we will now confirm, the ones who failed to send back their diaries. But we'll get more of them with PPM. It involves a lot less work. And every person of that type, who doesn't really use radio much, but is now in the survey, with their behavior accurately recorded, will make radio look weaker.

Conclusion? Potential problems for the sales department aside, how can you not be in favor of a more accurate rating system? Personally, I'm hoping that the minute by minute data will help us do a much better job of managing how we put together our programming, schedule our commercial breaks and sequence our music. Those are new opportunities that we should all be excited about. Agree?