We’ve spoken to dozens of programming people who are absolutely convinced that changes in PPM minute-by-minute meter counts provide useful actionable information with which they can make more informed programming decisions.
Yes, it would be great if we could use minute-by minute to fine-tune our programming.
Unfortunately we can’t, despite Nielsen’s assurances.
A decline in meter counts does not prove listeners dislike what they are hearing. An increase doesn’t mean they like what they are hearing.
The unfortunate truth is that minute-by-minute is astrology for programmers.
Those who believe in astrology are absolutely convinced of its value and accuracy. The true believer will cite instance after instance where his or her astrologist was spot on.
So it is with programmers who trust minute-by-minute. They know its right because they’ve seen it with their own eyes.
Minute-by-minute provides just enough seemingly accurate information to assure radio people who want to believe that it works.
The reality is that PPM was never designed to be accurate at the minute level.
It was designed to provide monthly rating estimates with an acceptable margin of error, error on the order of what diary measurement produces.
That’s it.
It was never designed to produce accurate audience estimates over smaller increments of time.
As you look at successively smaller increments of time you are looking at more and more randomness and errors in the numbers.
Weeklies are less accurate than monthlies, daily numbers are less accurate than weeklies.
And most assuredly, it was not designed to be accurate at the minute level.
At the minute level the noise and randomness overwhelms what little useful information might be there.
Yet radio programmers want to believe it is accurate at the minute level. They want to believe that there is something useful there.
And Arbitron wasn't going to pass up the financial opportunity to exploit this vulnerability.
The chance to squeeze a little more money from radio was just too tempting, so minute-by-minute was born.
“But it works,” you say.
“I can find hits with it,” you say.
“I can flag bad bits with it,” you say.
A trap that decision-makers too often fall into is what’s called Confirmation Bias.
We tend to look for information that supports what we already believe.
Confirmation bias is why Conservatives mostly watch Fox News while Liberals watch MSNBC.
We also tend to unconsciously discount new information that’s contrary to our views.
Conservatives who watch MSNBC will tend to dismiss more of what they see there than what they see on Fox.
We are all wired this way.
Radio programmers have to make a conscientious effort to avoid falling into the trap of Confirmation Bias if we are to make better decisions.
And looking at what appears to be objective data like meter counts makes it all the more difficult to resist falling into the trap.
What do we find when we objectively look at large data-sets of minute-by-minute?
We’ll show you in a future post.
If you use minute-by-minute, be sure to watch for it.
Comments