Music Research: What is it Good for?

If you filled in “Huh! Yeah!” you’re remembering a single from 1970 or a Seinfeld episode from 1994. But the response in this case is not, “Absolutely nothing!” While there are lots of tools available to programmers today to help them gain insight into which new songs are going to be hits in the near future, well-screened, passively-recruited music research remains the one way to know which songs are hits with your station’s listeners right now.

Fans of a specific artist or genre may tag songs that pique their interest on Shazam – and many of the most-tagged songs turn out to be hits, eventually. But, you won’t know how your station’s listeners feel about these emerging titles without music testing.

We love actively-recruited tools like RateTheMusic for their promotional value. Airing announcements asking listeners to participate and having links on your website gives a public face to the care and effort you expend selecting music. But, the “sample” of hyperactive fans, music promoters and competing-station personnel won’t give you a reliable basis to make decisions. Compared to properly-conducted music research, the scores line up about half the time – and it’s impossible to predict which half.

MScore scrapes PPM data to line up songs with what meters did when the songs were playing. Optimally, you get an understanding of which songs cause listeners to switch away in search of a better song. Sometimes you learn which song happened to be on when she switched to another station to find a traffic report or when her mood changed because of the weather.

Gauging consensus requires a properly-screened, passively-collected, statistically-valid, representative sample of a station’s audience. It used to require callout, but now online sample and data collection have turned that into what we at NuVoodoo call Online Music Research: full-sized, screened, passively-collected samples giving reactions to all the titles in consideration for a station’s current playlist.

Callout made sense when the phone was the consistent initial point of contact for Arbitron respondents. But, Nielsen has moved away from the phone in the face of fewer and fewer people having landline phones. To try to minimize expense, some callout vendors ask clients to use rolled samples, collecting half a sample this week and combining it with the half-sample from the prior week.

Callout depends on the respondent giving their opinions about three dozen hooks they’ve heard over the phone for free at a time when they may be trying to do other things. In the online space NuVoodoo occupies, respondents listen to hooks on their computer or smartphone at a time that’s convenient for them in exchange for compensation (like Nielsen respondents).

At NuVoodoo we think our Online Music Research is a savvier, more valid approach to testing music than callout (and why we don’t offer callout). But, both Online Music Research and callout help programmers gauge consensus among their audience about all the titles in consideration for their playlist. Many years ago, music research was a secret weapon among a handful of radio stations. Today, it’s the ante needed just to stay in the game.