Legend has it…after radio was invented, the next task became determining where to place stop sets. And for as long as anyone can recall, that same question continues to be asked. While everyone agrees it does matter where stop sets are located maybe now we’re seeing why and where to place them. Arbitron shares some observations based on experimentation with PPM data.
Gary Marince
By Gary Marince
Legend has it . . . after radio was invented, the next task became determining where to place stop sets. And for as long as anyone can recall, that same question continues to be asked. I’ve heard very convincing arguments with support for all positions, but I thought it might be helpful to share some observations based on experimentation with PPM data. I hasten to add this is just an opinion. Maybe we could consider it a thought starter. I think everyone agrees it does matter where stop sets are located . . . but maybe now we’re seeing why and where to place them.
For starters, we might want to consider looking at dayparts in a more fluid manner. Here’s why, if you take a look at AQH Persons at the Quarter Hour level, you’ll see every market and every station has a unique “listening profile.” And every market and station has a “peak period” where listening levels are the highest overall. Again, this can be measured and illustrated wonderfully with PPM. Say for example, the market peaks between 6:45AM and 7:15AM. And let’s say our station peaks around the same time. Well, I’m wondering if we’re not inhibiting our AQH Persons number by running the same clock in the 6:00AM, 7:00AM and 8:00AM hours.
Here’s the thinking . . . if you look at ratings in their most basic form, the Quarter Hour, and for the moment pretend there is only one Quarter Hour we have to address . . . we can do some quick math. To keep it simple, let’s say we are a music station and we have 12 minutes of music and 3 minutes of commercials to program. We’ll go with conventional wisdom for now and make the assumption that when a commercial first airs, it can result in some degree of tune-away for the station. Since Arbitron and Coleman Insights have released studies suggesting the percentage of listening retained during an average stop set is about 90%, we’ll say we lose about 10% of our listeners, at least for the short term, when spots run. Let’s say we start our Quarter Hour with 30,000 Persons . . . and if we place the stop set at the beginning of that Quarter Hour, mathematically, we’ve lost about 3,000 listeners and we’d end up with an AQH Persons of about 27,000 . . . or ten percent less than we could have had. So, with spot placement at the front end of the Quarter Hour, in theory, we deliver about a 27,000 AQH Persons to our advertisers. But, again with the assumption we only have one Quarter Hour to worry about, if we run that same stop set, we would be delivering an AQH Persons of approximately 30,000 to the advertiser . . . everyone wins.
Of course what we do in one Quarter Hour will eventually impact the next Quarter Hour and so on but if we’re mindful to make certain our peak Quarter Hours, those for the market as well as the station, are “clean,” we have a better chance at increasing our overall AQH Persons.
And by no means am I suggesting anyone consider running four stop sets with three minutes of spots at the end of each Quarter Hour. I’m merely using this example to make the point that where spots are placed can dramatically impact Quarter Hours and whole dayparts. Mind you, for non-peaks listening times, this concept might not apply – hence the consideration for looking at dayparts in a less monolithic or homogenous manner.
On another topic, with the recent release of PD Advantage for PPM, there has been a lot of interest in the “Limited Report.” This report is designed to help PDs gain comfort with the PPM Panel approach. I’ve had a number of phone calls from PPM market PDs asking “how do I know if the changes in my ratings were caused by my programming or if they were the result in Panel changes.” This report helps answer that. It trends Average Quarter Hour Persons two ways; the first view is the Total Panel number. This is what everyone sees, it’s what’s reported in the Analysis Tool, it’s the currency number and it’s what’s used in the traditional share calculation. It includes all panelists whether they were in the panel for a single month, all of the months or just some of the months listed. This report is best viewed as a graph. While the first line of the graph is the Total Panelist’s line, the second line is the Limited Panelist’s line. The Limited Panelist’s line consists of ONLY panelists who have been in the Panel for at least 75% of the “time” – with “time” being defined as the survey’s listed or included on the graph.
What we most often see when viewing this report is the two graph lines follow the same trend; up and down. What this tells us is whether we look at the total panel or a subset of the panel . . . if our station is improving or softening, it’s “panel wide.”
The included graph helps illustrate this point. Through the graph you can see how this Philly station is trending over the three survey period . . . regardless whether we look at the Total Panelist graph or the graph based on panelists who have been in all three surveys, the direction is the same.
Hopefully this will help us gain even more confidence in the Panel as we transition to it.
That’s it for now, keep the questions coming and we hope to see you at CRS 40 and the NAB.
Gary Marince is Vice President of Programming Services and Development for Arbitron, Inc. He is available to answer your inquiries regarding the Arbitron PPM or Diary services. You can reach him at gary.marince@arbitron.com.
