One of my goals in regularly writing this column is education. As you know, I’ve been around the ratings game for too many years, and having worked at rating services and with group owners, I’ve talked with many market managers and program directors.
While some are ratings-savvy and want to learn all that they can, others are absolutely terrified of dealing with the estimates. Seeing numbers scares far too many people in the business, especially programmers. In my experience, market managers tend to be more comfortable with ratings, perhaps because many came through the sales side. Getting ahead in sales typically means working with numbers.
If you’re uncomfortable with ratings, I’m here to tell you that it’s really not hard. For years, I’ve said if you can add, subtract, multiply, and divide, you’re good. You were probably taught how to do that in third grade. And even if you don’t think you can handle those four activities, there is a calculator on your phone.
My iPhone has one, but I still use my KMC-7000 solar calculator, which celebrates its 40th birthday this fall. When I started my doctoral program in the fall of 1983, we were required to take research methods, which meant statistics. This being the ‘80s, we had to solve some of the problems by hand, so you needed a good calculator.
While that exercise would be unthinkable today, I drove to the KMart in Okemos, Michigan (just east of East Lansing and the Michigan State campus) and purchased the KMC-7000. 40 years later, I still use it because, well, it works and it’s easier to use than the iPhone’s version.
In order to analyze and understand ratings data, you don’t need any sort of advanced math. Calculus might be involved in a reach and frequency model, but that’s why mathematicians exist. They design the models, while you enter numbers and hit a button. For programming analyses, it’s all simple stuff.
I’m convinced part of the issue is that programmers earned their jobs for a different skill set. Sitting in the PD chair, they’re expected to have a knowledge base that possibly wasn’t required, but once you’re in a leadership position, it’s hard to admit that you don’t know something. Just in case, let’s start with the basics.
If someone stopped you right now and asked the definitions of rating, share, cume, AQH, and TSL, could you answer correctly? Maybe? Let’s define the terms here.
RATING: The number of people who listened to a station, stations, or radio in general divided by the population in any daypart. For rating, the base (denominator) is everyone, whether they used radio at that time or not, and the base doesn’t vary regardless of what daypart you run.
Nielsen Audio updates the population estimates each year (October monthly for PPM and Fall for diary markets). The number is multiplied by 100 (same for share and cume). Typically, ratings numbers are very small, typically just a few tenths at best for any one station. FYI-television’s base is different and does not include everyone.
SHARE: The number of people who listened to a station or stations divided by the number of people listening to radio at that time. For share, the base (denominator) varies for every analysis, but think of share as your piece of the radio pie. And again, the result is multiplied by 100.
CUME: The number of different people who listened to a station, stations, or radio during a daypart. This is the reach estimate and everyone is equal, whether they listened for one quarter hour or one hundred. Cume can be reported in persons or as a rating (once again, the latter is multiplied by 100).
AVERAGE QUARTER HOUR: I covered the definition of a quarter-hour a few weeks ago. We know that the size of the audience varies depending on the time of day. For example, in PPM, the audience for most stations is larger in the afternoon than the morning (PPM turned the old listening rule of thumb on its head).
However, if you look at a Monday-Friday, 6 AM-7 PM run, the results for each quarter hour are different and are averaged across the number of quarter hours in the daypart, in other words, an average quarter hour.
TIME SPENT LISTENING: It’s a simple equation of quarter hours in the daypart times the AQH persons estimate. Take that number and divide it by the cume persons estimate. That will give you TSL in quarter hours.
This is just a starting point. There are any number of ways to analyze ratings data, but you need a foundation which means being fluent in the terminology. Arbitron used to have a booklet called “Terms For the Trade” and Nielsen Audio also has a similar one sheet (you can ask Google to find them for you).
If you’re not comfortable memorizing this stuff, find the documents, save the files, and even print them.
With the definitions out of the way, we’ll move on to some fun ways to parse the data in future columns. OK, fun for me, but valuable for you.
Let’s meet again next week.
