In the last two columns, I’ve suggested that the radio/audio audience measurement system from Nielsen is outdated. I’ve covered some of the data collection issues and made suggestions. This week, I’ll offer some thoughts about metrics.
Keep in mind the goals for an updated system:
· Simplicity, meaning fewer metrics but not necessarily a simpler collection system—in other words, simpler for transacting business
· Incorporating multiple data sources and “big data”
· Eliminating the diary service and using PPM everywhere
· Most important: a less costly system
Before you think that I’m suggesting we tell Nielsen Audio to get lost, I’m not. Nielsen is likely to be with us for many years to come. If another company wants to challenge them, so be it and welcome to our world, but this is not a “F— Nielsen” screed.
Now that I’ve offered a template for data collection, what metric should be reported and become the new radio currency? The advertising world is moving to “impressions” (you could argue that I should have used the past tense of “moving”), a term I’ve never fully understood because I thought that’s what we were selling all along. Selling access to an audience of an agreed-upon size based on estimates supplied by a trusted neutral third party creates impressions. As someone who grew up in a Kodak family, that’s like calling a “picture” an “image.” Perhaps you say “12” and I say “a dozen.”
Last week, I buried the average quarter hour measure (AQH). The funeral was lovely, with lots of tributes from radio people. You should have been there—or at least you could have sent flowers. But if we don’t trade on AQH, then what?
The key word in AQH is “average.” When you look at the AQH for any station for a daypart/demo, it’s an average of the estimated audience for all the quarter hours (QHs) in that daypart. Some QHs have more audience and some less, but in nearly all cases, the advertiser is buying the average, not the audience in the quarter hour in which the spot ran. For example, a spot in morning drive is bought and sold based on an AQH estimate for morning drive. If the spot runs at 6:50 AM, it reaches one size of audience. A spot at 8:20 AM reaches a different size audience.
Cume has similar properties. It’s the number of different people in the desired demo that can be reached during that daypart. You won’t reach them all unless you buy one spot in every quarter hour of that daypart, which entails a great deal of wasted spend on the advertiser’s behalf as well as a lousy experience for a long listener.
My suggestion is to make cume the de facto metric and report cume for every quarter hour throughout the day. Dayparts will look exactly as they do today but without the AQH metric. Nielsen and/or a competitor can report cumes by daypart as well, just as is done today.
The next question is qualification. Today, PPM requires three non-consecutive minutes within the clock quarter hour, and the diary service requires five consecutive minutes within the clock QH. My suggestion is that any minute within the quarter hour counts toward the quarter hour cume estimate. Yes, one minute gets you the quarter hour cume estimate.
Your first reaction might be, “Huh? You want to turn one minute into fifteen?” Any estimate from Nielsen, Eastlan, or any other source is just that—an estimate. A spot runs at a given time, and the audience varies every single minute—and it’s probably fair to say every single second. That’s before we account for the large amounts of sampling and non-sampling error involved, which we rarely do. Like all estimates, these are guides and offer a basis for buyer and seller to agree on pricing.
While my vision may seem cloudy to some of you, this is a system that uses the PPM panels in place today combined with streaming data (Nielsen can do it or license it from Triton), DTS AutoStage (again, licensing), and potentially other data (cell phone sample through apps, etc.). Next, much of that data would be “adjusted” using the PPM panel to “calibrate” the big data. It’s not easy to do. Nielsen has jumped through a lot of hoops to get their big data plus streaming system into the video service, but there are smart people out there who can figure it out.
Is this a perfect solution? Of course not. Some formats would likely argue that their appeal is for longer listening and that there is an upward bound on the format. For example, any Spanish-language format is likely limited to people who speak Spanish. It’s tough to build cume beyond that point. The same goes for urban formats, while CHR has always been a cume format with lower TSL. That’s mostly true, but those “limited” formats also “super serve” an audience that isn’t easily reached any other way.
Initially, I thought I’d be done with putting my thoughts out there this week, but there’s more to say. For example, in a cume-only world, how do programmers evaluate and make adjustments? Would the proposed changes eliminate certain formats as PPM did to Smooth Jazz? And most importantly, how do we move forward to a new reality, whatever it may be?
Let’s meet again next week.
Barrett Media produces daily content on the music, news, and sports media industries. To stay updated, sign up for our newsletters and get the latest information delivered straight to your inbox.
