When PPM started, the radio industry dealt with two huge changes compared to the diary service. Cumes generally skyrocketed while AQH audiences declined. This should not have been surprising because the diary service requires that respondents know the stations they “hear” (and “hear” versus “listen” spawned a minor controversy in the ‘80s) in order to report the listening.
With PPM, that requirement doesn’t exist and as a study some years ago showed, many panelists heard stations that they didn’t know they listened to, sometimes background sound, and other times, it was a station chosen by someone else, perhaps at work.
AQH declined because the diary service allows a respondent to draw a line to denote long listening periods and in other cases, the line isn’t even needed. In PPM, actual behavior is measured. Can you work an eight-hour shift without a visit to the bathroom? Doesn’t federal labor law require breaks over the course of the workday? You’ve never gone to a meeting or taken or made a phone call at work? PPM catches those gaps in listening.
Then one other “truth” came out from PPM. The typical listening occasion was claimed to be around nine to ten minutes. Well, yes, that can be true, but it is also misleading. As I noted some weeks back, there are edit rules for PPM and diary and these are typically complicated. For example, one of the PPM edit rules concerns the “lead in” edit which can credit one more minute to a listening occasion.
There is another edit rule called a “bridging edit”. Here’s how it works: If you are listening to station A, in other words, the meter is picking up station A’s code and then a gap of one minute or less happens (not some other encoded audio), and then station A’s code appears again, the “no code” minute is credited to station A. Makes sense that the panelist was probably listening to station A during that minute.
Now, let’s start with station A again. The meter picks up the code and then there are two minutes with no code. After the two minutes, station A’s code appears again. The result? One occasion has ended and another one has begun.
When I worked for Cumulus, I’d occasionally do a presentation for talent about how ratings worked. One purpose was to explain why your PD or OM was not crazy when he/she asked you to do certain things, but it was a matter of “beating the system”. That didn’t mean the PD wasn’t certifiably crazy for other reasons, but the diagnosis of his/her mental health would have nothing to do with a station’s formatics.
I also showed an example of how “occasions” can work. Pre-COVID, most working people commuted to work. Let’s consider that a panelist leaves home at 7:30 and turns on his/her favorite morning show. For the purposes of this example, don’t think about tuneout due to long stop sets, etc.
After starting out, the panelist stopped at a drive-thru to get coffee, and the radio’s volume was turned down for more than a minute. He/she continued on the way to work but then took a call for two minutes. After the call, it’s back to that favorite morning show. The panelist took 30 minutes to get to work.
Sounds like nearly a half hour of listening and at least two quarter hours of credit. It is, but this listening “session” included three “occasions”, specifically, home to coffee stop, coffee stop to phone call, phone call to work. The average will be around nine minutes, but the “real” listening session was a half hour.
This is not to say that some listeners still switch stations on a regular basis, for example, whenever a stop set starts because they’ve been trained to expect commercials and generally lots of them. Also, the 9-10 minute average may include occasions that are shorter than five minutes. Those occasions don’t end up in the ratings because as we all know, five minutes within a clock quarter hour equals a credited quarter hour. Everything else doesn’t matter. Nonetheless, PPM records all of these occasions and they may have been included in the studies.
What does all this mean for you? Most of us have shorter attention spans, a result of the myriad changes in the volume of stimuli over the past number of years (mine is gnat-like, just ask my girlfriend if you want confirmation).
However, many listeners can and do listen longer. In the Arbitron days, there was a long-ago claim that the average listening duration in the diary service was around 90 minutes. While I have little doubt that a contemporary study would lower that number quite a bit, in PPM, my bet is that the average “session” is closer to 30 minutes when you take account of the edit rules.
Again, whether you go with the 9-10 minute number or 30 minutes, both are averages. If you’re in a PPM market, use the Analysis Tool as I recently suggested. Dig into the data to find your strengths and weaknesses as well as those of your competition.
How do you use this to your advantage? I don’t want to contradict what many programmers and consultants have been doing for the last 15 years or so, but consider that much of your audience may be listening longer than you think. All of the tenets of good programming still apply, but if you believe that it’s “ten minutes and goodbye”, it’s probably not the case.
Let’s meet again next week.
