By Nick Holmstén
Photo Credit: NRK P3
What is the value of a song?
For many pop artists, it seems the answer is “a lot.” News broke last week that Justin Bieber is hoping to own the master recordings of any future music he creates. Meanwhile, Taylor Swift after struggling to buy the rights to her original work decided to re-record those hits so she could own the newer masters. In parallel, savvy businessmen like Merck Mercuriadis, and his company Hipgnosis, are literally buying up huge catalogs of music.
Why? Because they all believe that a song’s value is forever. But, is a song’s value really forever? Today, with streaming revenue now dwarfing physical sales (see RIAA 2020 report), the first week of album sales no longer forecasts an artist’s revenues like it once did. Instead, the value of an artist’s music is disproportionately weighted to the tentpole marketing strategy used to drive streams upon release — and the impact is long term.
Beyoncé’s streaming success is a great example. In 2013, in advance of the launch of her self-titled album, her team shifted their marketing strategy to invest heavily in a splashy digital campaign to catalyze the first week of listening. Mark Mulligan, a well-known and regarded music analyst, wrote at the time that because of this shift her first week of sales increased by 89%, reversing a trend of significant declines in sales for her previous albums following Dangerously in Love.
But Mulligan’s assessment of Beyoncé’s success still focused on the metrics of physical sales. Now with hindsight, we see that the true value was its long-term impact on her listenership overall. The self-titled album went on to cumulatively net a billion streams, a first for her; and its success propelled further listens of her previous (and as Mulligan notes, less successful) album, I am… Sasha Fierce which became her second album to receive a billion streams — and that momentum carried over into her subsequent 2011 album 4 that also reached a billion streams. And here is where we’re getting at the real success of Beyoncé’s digital campaign: it told Spotify’s machines that Beyoncé was an artist to promote. It gave machines reasons to promote her music across the platform’s ecosystem because there were ample positive user signals.
Streaming services are designed very similarly to social networks and other content recommendation platforms. On Spotify, “popularity,” “recency” combined with users’ behavior (“saves,” “add to playlists,” “replay,” etc.) are some of the most important indicators of the popularity of a track within that specific taste/genre.
Because consumers’ attention spans are fragmented and fleeting — these services reward creators of successful content.
This leads me to my next important point: consumption on streaming services like Spotify is increasingly more lean-back, driven by playlist listening — that is driving increased long-term value of songs.
When I joined Spotify in 2013, after they acquired my playlist-focused startup Tunigo, most listening on Spotify was done via a search and play application. Listeners would collect songs they liked and add them to their personalized playlist or library. While machine learning was already a massive priority for Spotify that was informing recommendation-based listening, my Tunigo colleagues and I had a different perspective on how to increase overall listening.
Our strategy was to curate playlists for all kinds of moments in life and build strong brands for each genre like ‘Today’s Top Hits,’ ‘RapCaviar’ and ‘Viva Latino’, combining human editorial expertise with Spotify data. During the seven years, I spent at Spotify, the percentage of consumption from Spotify playlists increased rapidly year over year and is way over 30% today and continues to grow.
So, as it relates to Beyoncé’s album, the massive marketing push to mobilize her ‘Beyhive’ to drive listening told Spotify’s algorithm to promote the music across the platform to new audiences. If other users liked similar tracks (good article on this topic below), there is a fairly big chance that Spotify would recommend a track through a personalized recommendation, an editorial playlist, or a combination of both. Before digital platforms created customized radio-like playlists for listeners, a song’s shelf-life directly correlated with airtime on traditional distribution channels (radio and broadcast television). But now, because algorithms drive curation on streaming services — a song’s inherent value extends and can cumulatively be considerably more valuable than previous measures of physical sales.
While at Spotify I continued to try and understand the limits of the success of a huge digital campaign. In 2018, we partnered with Drake to create a first-of-its-kind promotional takeover campaign for his album Scorpion. From custom playlists to full-page banners throughout the site and app, Drake was everywhere on the platform. Unsurprisingly the tactics netted major results, making the album the first ever to receive a billion streams in a week. The more interesting learning was how that initial push impacted listenership for his album over time (like Beyoncé). Given the listenership peak of Scorpion was so dramatically higher than any other of his albums, one might have assumed a dramatic drop in streams after the campaign ended.
But instead, the decline was gradual like previous releases and as a result, the total streams at each marker — one month, six months, or one year from release — were all significantly higher than previous releases at those same markers. Why? Because after the campaign subsided the Spotify algorithms had already determined via all the positive and reinforcing data points that Drake’s music was worth including across a range of playlists, giving the music a lifespan that outlasted the original campaign. Cumulatively, Drake’s listenership for Scorpion remains the highest across all of his albums on Spotify. And here is why a song’s value is forever: if artists can make a large enough digital push at release, they are able to create a long-term listenership for their music that never existed previously.
The first week of album sales is no longer the indicator of success, but that doesn’t mean the first week of marketing isn’t as important. It’s actually more important than ever.
Tentpole marketing campaigns around the launch of an album have the ability to influence digital algorithms — which in turn means that the music’s shelf-life will be exponentially longer. Folks like Merck Mercuriadis get this and are betting big on it. And the reality is this paradigm is not limited to just music and includes film, TV Shows, etc. Any content network from Netflix to Prime Video has similar types of algorithms that will surface the most popular content because that hooks subscribers. This is great news for any creator who is able to successfully compete in the Attention Economy — and why it makes a ton of sense that the Biebers and Swifts of the world are actively seeking the rights to their master recordings.