A No-Nonsense Marketing Smart Tip January 19, 2005
The doorbell rings. It’s a new artist. He has a new song. “Welcome,” the research respondent says. She hasn’t heard of him or his music before. While his band sets up in the living room, he shows her his music video and tells her about his appearance on SNL and which episode of The O.C. featured his new song. He shows her his reviews, all good, from Entertainment Weekly and Rolling Stone. Bono told him he was one of his favorite new artists, she learns. He tells her about the small clubs he’s selling out and the enthusiasm of the fans who approach him after gigs. The band’s ready, he settles in, picks up a guitar and sings his new song. “How do you like it?” he asks with a charming smile. “I love it,” she replies. “I rate it very highly.”
Who is this artist and what is this song? We’ll never know. Because he never came to her door, he never showed her his video or regaled her with the story of his TV appearances. She never saw his reviews or heard about the sold-out shows and enthusiastic fans. She doesn’t know he’s Bono’s new favorite. And she has never seen him smile. Instead, she listened to an extended hook. It sounded weird. “I hate it,” she said, and she rated it poorly. The song sank unnoticed and the artist along with it.
The Pre-Testing Puzzle
What happens when you test new songs for an audience which isn’t the least bit savvy about music – Joe or Jane-average? They reject what’s unusual in favor of what’s comfortable and familiar. A hook which sounds like everything else they know scores well, anything else does not. The result of this exercise repeated over and over is evident across the Radio dial: The safe and familiar flourishes and what’s challenging at first impression vanishes. It’s a world without an Aeron chair, which was once considered “ugly” and is now stylish and popular. It’s a world without Seinfeld, because for its entire first season no one wanted to see a show about nothing.
Listeners vote on our music and then complain about it and flee to Satellite Radio or iPods – they want us to challenge them but they lack the means of telling us exactly how. Our music research doesn’t allow it. We try to test new material on them, but when that new material sounds odd or simply different from what they know, they don’t like it. Would it grow on them with experience? Would they warm to it the way they warmed to Seinfeld?
The Hook or the Hit?
I’m arguing that you can’t pre-test new music the way you test the familiar stuff, you can’t pre-test it with just anybody, and you can’t pre-test it without considering the context in which that music lives. To do otherwise is to predict the hook, not the hit.
Understand this: Hits and their hitmakers live in a media-rich world. They are everywhere the music-fan turns. They drop in on TRL. Their videos are in heavy rotation. They have a menu of CD’s and magazine covers. They are known and recognized and famous. Maybe they’re actor/singers or singer/actors. Maybe they’re tabloid fodder. It is into this maelstrom of context a new song falls. The minute it becomes a hit, it joins this three-ring circus. And where the song ends and the circus begins, who can say? One and the other are the same.
Picking the Hits the Smart Way
Allowing a random selection of listeners to rate brand-new music may be good research in that it correctly samples the population, but it’s horrible research in that the average person is not qualified to evaluate the hit-potential of brand-new music. When first impressions are so important, the sample should be made up of folks who are interested, not compelled, and certainly not bribed. As Malcolm Gladwell (The Tipping Point) writes in his remarkable new book Blink, regular folks aren’t music “experts,” and “to force them to be – to ask too much of them – is to render their reactions useless.”
Conveniently, these music “experts” are more likely to be found online volunteering their efforts for sites like Promosquad and their like. Music fans are more interested and passionate and thus better at correctly interpreting their first impressions, suggests Gladwell, and first impressions are exactly what pre-testing songs is all about.
Further, we need to provide more information – more context – for the music being rated because this is exactly the context the song will possess when it’s a hit. What does the band look like? How do my friends rate them? Can I see their video? Are they on MTV or Radio already? Can I hear other songs by them? What do the critics say? How are their club dates selling, etc?
Doesn’t this bias respondents? You bet. But where is the line between the song and the elements which surround it, between accuracy and bias, between the music and the package? In fact, they are one and the same. They never exist without each other – except in the arcane world of music research. Remember, without the “package,” Pepsi beats coke in taste-tests by a wide margin. But slap on the labels and the results are reversed. Research shows that the very same peaches taste better in a glass jar than in a can, and the very same ice cream tastes better in a cylindrical container rather than a rectangular one, writes Gladwell.
The song and the package are one. Pre-test a part without the whole and you’re pre-testing the hook, not the hit.