Smiling While Scratching My Head

If you, like me, are glued to the stats and musings of Richard Harker & Glenda Shrader Bos on Harker Research's "Radio Insights" blog, you are already up to speed on their recent format comparisons, which they feel is a more accurate use of the numbers than Arbitron's own "Radio Today" analysis.

Devoting the entire front page to Country’s performance,“the most trusted news in radio” declared. The format ranked first among all formats in diary markets with a 14.5 but finished fourth in PPM markets with a 7.1. Put another way, Country’s average diary share is more than double its PPM share. The statement is factually accurate, the ciphering flawless, but in the end meaningless. A few lines later we learn the purpose of this expose' when the author asks rhetorically, “Why do some formats perform better under one ratings system than under another?” It's perhaps a fair question, but using Radio Today Country numbers suggests that the author really doesn't understand how to go about answering the question. Radio Today shares are calculated just like you’d do for a single market. Share is just a format’s AQH divided by the market’s AQH. To calculate a national format share, add the AQH of all the stations, then divide by the total AQH of all the markets. Simple, right? But what happens when there’s no full market Country station in New York or Los Angeles? The two largest markets in the country don’t contribute any AQH to the format tally. Consequently, the format's share is much lower. Radio Today is not designed to compare formats the way the author has tried to do.

Both approaches make country look very good, fortunately.

For example, the Harker researchers found: "There are 90 Country stations in PPM markets, but only 60 have a full 12 months of PPM numbers, so like the other formats, it will take several more months before we know how the entire stable of Country stations are doing.

"That said, PPM appears to have been very kind to Country. Nearly two-thirds (63.3%) are higher today than they were a year ago. Country has the highest percentage of gainers of any format we’ve looked at. The median Country station has a 4.1 share 6+, a tie with Hot AC for third among the formats we’ve discussed so far. The big difference between Country and Hot AC is that Country has positive momentum while Hot AC is stalled, so in this tie we give the nod to Country."

As I reported in this space when "Radio Today" released, Arbitron's research team pointed out:

The No. 1 format overall, in both Diary and non-Metro areas, Country reached more than 65 million listeners each week on more than 1,700 stations—the most outlets of any music format, not including 300 Classic Country stations. It also was the top format among most age groups (including a big lead with adults 25-54) and in most dayparts. Country enjoyed some of the longest time spent listening of all formats, and its listeners set a new high in college degree attainment in Fall 2010.

Harker distrusts ARB's approach on statistical grounds and so to assess the health of each format they looked at two key metrics, median station performance and momentum. The median station is the one in the middle. Half of the stations in the format have a larger share, half have a smaller share. As they see it, the median tells us how the format is trending. "If we see the median rising, it means mid-pack stations, the majority in the format, are gaining share. Tracking the median station is more useful than calculating an average share because averages can be distorted by what are called outliers, stations that dramatically outperform other stations in the format."

New PPM markets coming on-line make it difficult to compare current median shares to last year’s numbers because the newest markets tend to be smaller than the first PPM markets. So, to look at median share trends, Harker calculated median shares for just the stations that have been in the format and measured by PPM for at least a year.

While the median shares tell how popular a format is, Harker posits that a more useful metric for indicating the future prospects of a format is momentum, comparing past performance to the present.

Harker calculates momentum by comparing each station’s share today to what it was last year at this time, using a three month average. Positive format momentum means that more stations are higher today than they were a year ago.

Why is it that there's such a big gap when ranking format mo' and yet so much compression in the 6+ median share ranker, where just a tenth of a point separates #1 and #2, back just .4-.5 are the next four and then just a share behind #1 is #7, leading #8 by just .2?

Richard and Glenda have a theory on that as well, in a 2010 post called Arbitron PPM Trends vs. a Coin Toss, most of the month to month movement of PPM is random.

Stations go up and down for no other reason than luck of the draw. Share estimates always have two components. The first is the actual number, the share you would find if you could ask everybody what they were listening to, instead of just a small number of panelists. The second component is statistical noise. It is interference like static drowning out the signal from a distant AM station. The noise is completely random and unpredictable. Sometimes it adds to the actual share, inflating it. Sometimes it subtracts from the actual share, deflating it. Both monthly PPM share estimates and quarterly diary share estimates have so much statistical noise in them that it drowns out genuine changes in month to month ratings.

I'd interpret that powerful momentum to country's high level of engagement, in spite of lower levels of cume in PPM compared to the formats above us in median 6+ share. And, putting it another way, as Arbitron reinforced in their diary/PPM data for the past year: " Country enjoyed some of the longest time spent listening of all formats." Harker would likely predict that the country format could be a rank or two higher in next year's median averages with that solid improvement (as long as we can keep it up).

There will be a lot to discuss when we're all together next month in Baltimore, which is why both Mike O'Malley and I will be there.

Meanwhile, it's nice to know that no matter how you count it, in PPM, diaries, by momentum over last year, average OR median shares, the country radio format is having one very good 2011.