Lando Griffin said:
1. The prime goal of a radio station (as a business) is to make money by providing advertisers with an audience that has and will spend a good amount of disposable income...
Ratings driven advertising buys are primarily based on a couple of factors which are age and gender (based on a product's highest usage groups) and station rank and cost of delivering each listener (CPP, CPM, etc). Since the US has very few people who are not consumers of most of the things advertised on radio, income data is really a secondary concern as nearly every station delivers some upper income listeners.
2. ...and the idea is to program to that large group of listeners who work and make that spending money...
Very few stations design formats to specifically appeal to upper income households or persons. We design stations to get as high in rank in sales demographics as possible.
3... and these "tests" take a certain amount of hours out of a day from people who are somehow not busy enought to take part in a long survey
Including driving to and from, most take about 3 hours total. For this, participants, depending on where they are, may get as much as $150. That is a nice extra spiff that many will see as, "now I can replace the broken blender" or " I can take the family to dinner this weekend" or somesuch use.
-- deducing that these people likely are not working or they don't have salaried-enough work that needs them to the point where they couldn't participate in such a long session...
Almost all music tests are done in the evening, after most conventional workdays have ended. Or they are done on weekends. They fit into most people's schedules quite easily, and, as I said, provide excellent compensation. Sessions usually include snacks or a buffet, too.
4. ...thereby leading a thinking PD to believe that this "group" that determines what "tests well" is not a very likely sample of the demographics, psychographics AND invome levels they are looking for...
A recruit for a music test for an existing station is almost always based on a cross section of users, mostly P1 and P2, who listen enough hours to know the music well. If there is a direct or flanking competitor, some of those listners may come in. Recruiting achieves a balance in gender and age that reflect the composition of the core audience for the station.
Things like income, if the recruit is random, will be self-balancing. You will usually get a cross section that represents the community.
They WHY should these same people be relied uipon to determine the playlists of radio stations?
Properly recruited, a music test will very closely parallel Arbitron respondents.
The "lowest common denominator" mihgt work well when selling minor products like Budweiser, but the advertisers who want a higher-income level of loyal listener may learn that this kind of program dictation would be irrelevant for the most part.
I do about 100 tests a year, and overall I have found that unemployed seldom get recruited. First, the US is near an historical low in unemployment, so there are not many. Second, we find when we do background data questions, that most work, and those that do not are women who by choice work at home taking care of a family or they are students. It pretty much matches the composition of the workforce, too, on white and blue collar jobs.
Case in point: A big company-owned radio station plays an AC format, yet has "Wonderful World" by Louie Armstrong in its regular rotation -- because it "tests well." (It does, however, sound funny when coming after Kelly Clarkson
Almost every listener has three or so favorite staitons they use regularly by personal choice (not ones they have to listen to at work, for example). So there will be songs that will test that don't belong on your station... most of us know not to test songs we would never play based on stationality. Some PDs are not that bright, or think this sort of idiocy adds variety. It's not good programming, in my opinion.
Just a question. You seem like the guy to ask, David.
Those are great questions and observations. Thanks for making the post.