How Spotify are using data to tell personalised stories about music taste...
Like a performance review of our leisure time, the Spotify Wrapped feature presents music consumer's year back to them in little snippets and narratives. What does this mean for music consumption?
Anyone clicking on Spotify this week will have been confronted with the 2020 Wrapped feature. Spotify usually provide users with their own unique end of year playlist, compiled from the tracks they have listened to most. The Wrapped feature adds an extra layer to this self-reflection, using harvested data to generate personalised stories about the user’s tastes and music consumption.
Picking out small snippets, these condensed narratives use the tip of the data iceberg. Despite only curating a few nuggets, it's still suggestive of all the analytics and pattern recognition that are going on in the background. It's intriguing to see what you've been listening to captured in neat little highlights, yet it also forces you to consider all that data that churns through these systems. There's nowhere to hide from your listening data - you can't pretend to like something cooler or something more obscure when your data says something different.
Spotify's Wrapped presents our identity to us in miniature automated tales, adding a few little evaluative nods along the way. I'm a "pioneer" it tells me, because I listened to D Double E's track Contact Us before it reached 50,000 plays. I think Spotify's algorithms may be humouring me. A bit of ego boosting to keep me on the platform. The tone of the crew neck capitalist permeates these snippets.
It also tells me that I've listened to 210 different genres. I suspect the task of writing a full list of these 210 music genres would be a difficult task for anyone. Spotify don't provide a list beyond the top 5, so I'm left guessing what the others might be. The presence of the mysterious label “Chamber Psych” at number 3 and bookended between the more recognisable Alternative Rock and Grime hints at the problem. What this does suggest is that in the presence of music streaming, genres have fractured into tiny little shards. Working through into the capillaries of genre labels the differences between them must be miniscule and, I suspect, hard to identify. Perhaps the self-organising presence of the streaming playlist is in part driving this level of genre labelling. Or perhaps it is Spotify trying to make connections between the listening practices of its big population of streamers that requires all these categories to be formed. More categories mean a greater level of granularity when comparing music taste.
It is strange to see music taste presented in this cumulative and stark way. In some ways it feels like a performance review of my leisure time. I'd listened to 349 new artists this year, apparently. Spotify informs me that this shows that I've "found ways to grow". More boosting to keep me coming back perhaps.
It's the top 5 rundown that, like the old chart shows, is the big draw. It turned out that the Sleaford Mods were my top artist. Their track Tweet Tweet Tweet was also the top track (the song that, they say, “helped get me through it all”). I also discovered, with a bit of surprise, that I was in the top 0.1% of Sleaford Mods streamers. That was a different sort of metric, as it was directly comparing my listening to other streamers and evaluating my listening in relation to them. This made it feel more like it had become a competition in which we were competing to be the most dedicated fans. It was like the data equivalent of who has the most well-worn band merchandise, the most limited edition singles, and encyclopaedic knowledge of lyrics or the quickest recall of band facts.
This could be seen as another way that tech platforms are presenting our near past back to us in packaged form. Social media are cluttered with reminders of past moments. Spotify's Wrapped feature is part of that trend - in this case it uses data to present our recent musical biographies back to us. The next logical step will be to get notifications telling us what we were listening to on this day five years ago etc.
Inevitably Spotify’s platform stickiness is also confirmed. My story revealed that I’d spent 20,466 minutes streaming music on there in 2020. I worked out that that's 341 hours. Over two weeks of constant listening. That's 3.89% of 2020 that I spent on Spotify. I'm sure others will have spent even more time on there. No one ever knew how often a CD or record span, but streaming is a constant source of music data.
By using data to confront us with our tastes, maybe the Wrapped feature will change how we listen to music in the future. Wanting to change how our music taste is narrated, we might start 2021 by trying to be more eclectic, by forcing ourselves to listen more, we might put cooler tracks on repeat and resist the temptation to play old favourites or something we might worry looks a bit naff or lightweight. The stories that can be told about our music tastes might change as we glimpse into these snippets of data.
Like putting our music on display in racks and stacks (you are encouraged to share your Wrapped results with others), Spotify are finding ways to make our streaming visible. When we know we are being measured we are likely to change our behaviours and be reactive, that's the power of metrics. Spotify's algorithms already quietly guide our choices, now this use of our data to storytell our tastes may mean that we will try to shape our listening habits to reinforce the version of our identity that we would like to realise.
With next year’s results in mind, I'm just going to put on another playlist to boost my listening minutes and enhance both my artist count and genre score.
If you are not already a subscriber and you would like to receive future pieces by email, you can sign up here…