How Small-Market Stations Can Optimize Nielsen Ratings Despite Smaller Samples

When I’ve passed along some of my tips and tricks for interpreting Nielsen estimates, the spiel has generally been about either PPM markets or the larger diary markets. This week, I want to take a look at the “condensed” markets better known as the “two-book average” markets, in other words, the smallest markets that Nielsen measures.

If you’re in one of these markets and don’t particularly like the averaging system, it’s understandable. However, I’ll defend Nielsen in this case because it’s a cost/benefit situation.  As the measurement system for smaller markets (the diary) has been stretched to its limits, the other part has been the (reasonable) reluctance on the part of broadcasters to pay more for the service. 

My understanding is that there are fewer agency buys than ever in these markets, so Nielsen’s role as currency, the reason that you pay so much for the service, diminishes. Two-book averaging brings a better chance at consistency. 

When I’ve dealt with two-book average markets, I’ve always wanted to look at the individual halves of the survey. Nielsen doesn’t make it easy to break out the components, but you can do it. Don’t try PD Advantage because the answers aren’t there. That doesn’t mean PD Advantage doesn’t have value in small markets, but the small number of diaries means even with the full sample, trends will get a little strange.

If you want to see what half of a two-book average survey looks like, you need to dig into Tapscan. Tapscan is a pretty straightforward piece of software and it’s not hard for you to find each half of your survey.

Go into the “Research” area of Tapscan and after you’ve done a ranker run to “wake up” the software, (just do the default for your market), click on the “Survey” option in the upper left corner. You’ll see that Tapscan defaults to the most recent combined survey (a Spring and a Fall or vice versa), but you can remove either “half” if you want. Just double-click on the survey you don’t want and it will go away. 

You’ll also get a pop-up that asks if you want to continue because you’re using a single survey which is less than the standard one-year survey. Click “OK” and you can see one-half of any survey that is available to you.  You can run all the standard demos and dayparts.

While this system works, it’s still limited. You can’t do trends this way, so if you want to trend the data, you’ll be doing it one survey, or more appropriately, half survey, at a time. It’s a very slow process and in my Cumulus days, I could do an analysis on a major station in a PPM market far faster than for a condensed market. Still, you have access.

What’s the value of splitting the surveys? Looking at the most recent “half”, you can see how that part of the survey performed for your stations compared to the entire survey.  It’s simple: if you were doing better in the second half, wonderful! Do well in the next “half” and you’ll be in good shape. If not, a request to your preferred deity may be in order. Ask for some heavy diaries in the next survey.

One thing to look for when you split out the survey this way is outliers. The term is simple: gaining listening where you didn’t expect it. Some obvious examples: a CHR station pulling big 55+ numbers or any type of rock format with strong female numbers.  It’s a little bit of detective work, but outliers are just that. They may be here one time but are unlikely to repeat.

Splitting out the survey also allows you to get an idea about any changes you may have made between the two surveys that make up the most recent book. Did you switch up a morning show, add some new syndicated programming, or something else? Here’s your first opportunity to receive feedback.

One other tip for small markets: I typically use persons numbers because the estimates are so small. So many programmers are hooked on share, but in many of these markets, 100 persons (the smallest increment in Nielsen Audio) may be equivalent to one or two share points!  You’ll find plenty of examples of one of your stations between two or three share points ahead of the next station(s) down the list. 

Look at persons and the difference may only be 100 and could be based almost entirely on the Nielsen rounding conventions.  Go back to your high school (or earlier) arithmetic: 151 rounds up to 200 while 149 rounds down to 100.  Nielsen would report the first as 200 and the next as 100.  Quite a difference, but the real difference might be as few as two people!

Now that I’ve been writing this column for Jason for over a year, some readers may have questions about how Nielsen works or for that matter, how other services work. Please feel free to send any questions along to me. I’m happy to answer and perhaps you’ll give me an idea for a future column!

Let’s meet again next week.

Dr. Ed Cohen

One of the radio industry’s most respected researchers, Dr. Ed Cohen writes a weekly business column, heavily focused on ratings research for Barrett Media. His career experiences include serving as VP of Ratings and Research at Cumulus Media, occupying the role of VP of Measurement Innovation at Nielsen Audio, and its predecessor Arbitron. While with Arbitron, Cohen spent five years as the company’s President of Research Policy and Communication, and eight years as VP of Domestic Radio Research.

Dr. Ed has also held the title of Vice President of Research for iHeartMedia/Clear Channel, and held research positions for the National Association of Broadcasters and Birch/Scarborough Research. He enjoys hearing your thoughts so please feel free to reach him at doctoredresearch@gmail.com.