It's no secret that the music industry is in turmoil. The transition from analogue media to digital downloads to digital streaming over the last couple of decades has seen great shifts in the way the industry is structured. And it continues to evolve. Spotify's latest financial results reveal massive growth in listenership, yet at the same time increased losses. After nearly a decade of operation, and despite 90 million users, Spotify has yet to make a profit. I don't think streaming per se is bad at all, but clearly there's room for improvement. And with the rise and demise of a different service or app almost every month, the playing field continues to evolve.
The soundtrack to our lives
Ok, so the business side is a bit off. But what about the music itself? What about the songs that form the soundtrack to our lives? It's tempting to say that music today is rubbish compared to 'the good old days', but is that really true? After all, every generation says the music they grew up with was the best. Could it simply be nostalgia rearing its friendly head to interfere with our objective perceptions? To some extent, certainly.
Let's also consider the changes in the way music, particularly hit music is created. Teams of songwriters are more common now than they used to be, and, as is the case in any superstar industry, a relatively small number of them reap the majority of the success and rewards. Same goes for the artists themselves. But that doesn't necessarily make the music they create any better, does it? (I don't know, what do you think?)
There's a great tradition of songwriting stretching back decades (centuries if we move beyond the age of popular music), and it's produced some phenomenal pieces of cultural history. Far from bashing songwriters, I'd like to suggest an additional approach to enhance their craft which would at the same time help develop the industry as a whole.
Focus on the experience
My view is that in order to create the best, most engaging music possible, one has to understand how that music is experienced by the listener. Not just whether they 'like' or 'don't like' it, since studies have shown that self-reported preferences judgments do not accurately reflect how much a listener actually enjoyed a song. No, if one wants to create songs that keep listeners coming back again and again, it's important to dig a little deeper.
Now of course, there are some major initiatives moving forward with this. A whole host of music tech companies have emerged in recent years, all hungry to take advantage of the mass of data generated by the digital revolution. From song features to play counts to user profiles, there's a phenomenal amount of data out there. But still, these companies are approaching it from the wrong angle.
Not one of these companies (and feel free to correct me on this) considers how listeners actually experience music, how they perceive it, understand it, or respond to it (non-verbally). Play counts, user profiles and MIR-driven music features are all well and good, but they are broad characteristics that are limited in terms of actionable insights. In short, everyone's analysing the wrong features.
Take, for example, how limited most music recommendation systems are, or how poorly computational models can explain what makes a hit song. With apologies to my colleagues working in this field, they're pretty terrible. And it's because they're focusing on the wrong features. We don't just listen to music, we experience it, and it's that experience that we should be quantifying.
The great thing is that the same alternate approach I'm suggesting for these tech companies to make better use of their data is the same approach that songwriters and composers could use to create more engaging music for fans to listen to. I'm a firm believer that there's no such thing as a bad song. But there are great songs. And it's great songs we all remember. Understanding the listeners' experience, and creating music that offers a maximally immersive experience would give any songwriter looking to create a lasting piece of cultural history a firm leg-up.
By understanding the listener' experience, then, one can create music that is more engaging to listen to, music that is more likely to have long-term success. Combine this with a better, more profitable industry driven by the same fundamental approach, and it's a win-win.
And it all starts by understanding the listeners' experience.