How Ed Sheeran Broke The Charts

Unknown.jpegUnless you have been hiding under a stone on Mars this last few weeks you will have struggled not to hear or see some clip of Ed Sheeran one way or another. Atlantic Record’s carpet bombing market campaign has tipped Sheeran into global ubiquity. At the centre of this approach is a ‘be everywhere’ streaming strategy which saw Sheeran clock up over 68 million Spotify streams in 1 day (a record for any single artist). Though, the 1 billion views he clocked up for ‘Divide’ on YouTube shows where the real streaming audience of scale resides. But what makes Sheeran’s ‘Divide’ campaign stand out is what it has done to the charts. Or rather, the weaknesses in the charts that ‘Divide’ shines a light on.

What Role Should Streaming Era Charts Play?

As of March 13th, Ed Sheeran’s ‘Divide’ album accounts for 9 of the UK top 10 singles, while all of the 16 tracks on the album are in the top 20. If there was ever a sign that streaming is breaking the charts then this is it.

The writing has been on the wall for charts ever since the recorded music business decided to incorporate streams into them. Doing so was a perfectly understandable move but it is one that has incapacitated the charts. As we predicted back in 2014, incorporating streams into charts would fall over because the charts were being forced into trying to simultaneously measure sales trends and airplay. As I wrote 3 years ago: “try simultaneously [measuring airplay] with measuring sales and you end up with a diluted mish mash that does not do either job properly.”

Underpinning all of this is an existential industry debate over whether streaming is replacing retail or radio. In truth, of course it is replacing both, but which is it doing more? The answer to that determines the role charts should be trying to play. However, the answer looks very different depending on where you sit. If you are a record label you see streaming growing by 57% in 2016 to reach $5.4 billion. Streaming is indeed becoming the future of retail. But it is also how you break artists and releases now, therefore it is a bit of both. Go over to the artist side of the equation and streaming becomes a crucial tool for driving exposure and helping sell concert tickets. As Ed Sheeran himself said during his last album promo cycle, for him it is all about live. Indeed, for most successful artists, recorded music revenue is just a small part of the revenue mix. So at its most extreme, streaming is a marketing campaign that pays you instead of you paying for it.

Reach Or Engagement?

In the old charts model an Ed Sheeran super fan buying ‘Divide’ and playing it a hundred times in the first week would only show as one sale, and an album sale at that. There would be no impact on the singles chart. But in the current UK streaming charts, not only does that fan’s album listening now get counted in the singles charts (instead of just the album charts), the resulting 1,600 streams (16 tracks*100) become 160 chart placings (100 streams = 1 sale for singles charts). Consequently, the charts are conflating audience reach with audience engagement. It is the equivalent of Facebook merging Monthly Active Users and Daily Video Views into a single metric. It wouldn’t work for Facebook and it just doesn’t work for music.

A Fiendishly Difficult Problem To Fix

There is no doubt that ‘Divide’ is a fantastically successful and popular album, the problem is that because the charts are conflating sales with consumption we simply don’t know just how successful it really is. And that does a disservice to both Sheeran and his fans. Don’t get me wrong, I truly feel for the various charts organizations across the globe. This is a fiendishly difficult problem to fix, but the current solution just isn’t working. In all likelihood, a dynamic solution is going to be needed, one that has the flexibility to evolve as the streaming market and its industry role changes.

The Time May Have Come For A Separation Into 2 Charts

Ultimately the recorded music business needs to decide what it wants the charts to measure. In old parlance: sales versus airplay, in contemporary terms: reach versus engagement. One near term fix would be to only consider cached streams towards the charts (perhaps with a smaller deflator than the current 100). This would have the advantage of making the measure more reach focused rather than engagement led. It would also have the effect of reducing the impact on ‘push’ curated playlists, which depending on where you sit, can be either an entirely good thing or an entirely bad thing.

If such an approach was taken then some sort of purer engagement chart would need creating to sit alongside the main chart, one that weighted total streams alongside traditional radio. The argument for a streaming-led airplay chart is even stronger than revising the sales chart. With playlists now accounting for 58% of all streams (see MIDiA’s Streaming Music Healthcheck report for more) and curated playlists a third of those, streaming is becoming less about on-demand and more about lean back, radio-like experiences. Streaming is seemingly making radio programmers of the entire recorded music business. It is time for a chart that reflects this change.

‘Divide’ is an exceptional album in terms of commercial performance and audience reach, as is its impact on the charts. But in the latter respect, it is simply a trail blazer for the way in which big albums are going to play out on streaming. ‘Divide’ might not be the hair that breaks the camel’s back but it has certainly fractured it.

A Quick Note On Charts

Context: I spent a couple of years working in an Internet measurement company in the early 2000’s so I have some direct experience of measurement methodologies.

If streaming is changing the music industry then it is only logical that the way in which the market is measured should change too. We are in a transition phase where consumers are abandoning ownership for access, and also often swapping spending for free. It is a period of uncertainty and there is an understandable clamour to make sense of it with new approaches to measurement. Hence Billboard’s decision to incorporate streams into its album charts (with 1,500 streams equalling one ‘sale’) and other moves such as the UK’s Official Charts Company’s incorporation of streaming. But it is not just the methodologies of the charts that are changing, their entire purpose is shifting. In fact, before we reinvent charts we need to be clear about what purpose we want them to serve.

Sales charts have historically served a number of purposes:

  • Driving sales: securing a high 1st week chart position dramatically increases the chances of more sales because it unlocks opportunities such as (more) radio play, TV coverage, retailer promotion etc.
  • Driving awareness: by being in the charts music gets in front of consumers which matters because they typically exhibit herd like behaviour. Chart success can beget chart success.
  • Measuring popularity: charts give a snapshot of what is popular with music fans

Adding downloads was a very natural and entirely logical step for charts. Downloads after all are still a sale. So all of the objectives remained intact. But adding streaming changes things. You might be able to make a direct revenue comparison with sales (i.e. 1,500 streams = 1 sale) but that obscures a much more diverse consumption picture. The simple fact is that many streams are not a sales equivalent.

A subscriber listening to an album a few times through on a paid streaming tier can reasonably be considered analogous to an album purchase (in entirety in behavioural terms and in fraction in revenue terms). But a YouTube user that listens to a handful of tracks from an album just once cannot reasonably be considered anything like an album sale. Instead this is far more analogous to radio listening.

Using a revenue determined measure to define the link between streaming and sales changes the nature of a chart and in turn also what purposes it serves best. It becomes something that sits between sales and audience measurement. And, done carefully, that can be OK, but this needs to be recognized. A chart that includes free streams cannot though be considered a sales chart.

Capturing streaming activity is crucial but the issues are where and how. Radio play charts are a fantastic tool and help identify taste trends. This is where free streaming data most naturally sits, along with other inputs such as Shazams and also social data captured by the likes of Next Big Sound and musicmetric. In this increasingly complex consumption landscape a broad range of inputs are needed to paint an accurate measure of music fan interest and behaviour.

But try to do that simultaneously with measuring sales and you end up with a diluted mish mash that does not do either job properly.