I’ve been on the fence about Spotify for quite a few years, since it was first launched in 2008 when I was at university. I could never decide if it would be a great thing or a horrible thing for the music industry. As a music consumer, a music creator, and someone who has a decent grasp of the ins and outs of music royalties, I feel like I’m qualified to explore the subject in detail.
It seems like every other day there’s an article about how some famous artist has made some tiny amount of money from Spotify royalties. Yesterday there were two pieces of news that inspired me to think about this further… La Roux claiming she got £100 for the last quarter, and “All About That Bass” cowriter claiming £3700ish in total for that song (178 million streams). These articles are clearly written in a “Spotify is evil” way, so of course they don’t give you any basis for comparison. That’s what I’m going to do here, then we can make up our minds with all the facts of how royalties are broken down.
This will be a case study of a fictional top 10 song by a fictional major-label pop artist, based on one week of fictional airplay, streams and downloads. It’s based on average stats which have either been researched or are common knowledge, with educated guesses filling the gaps. The point is to compare how royalties are generated and split for different listening mediums.
NOTE: I’ll be talking mainly about UK royalties for UK artists. Each country is slightly different, and international artists sometimes get paid differently, so the numbers are variable but the concepts will probably always apply. Most of these figures are estimates anyone, but ones from official bodies so they should be fairly accurate, and of course numbers change month to month.
INCOME FROM MAINSTREAM RADIO
I’ll use Radio 1 as an example since a lot of data is easily accessible:
– BBC Radio 1 apparently averages 9.7million listeners in primetime.
– A-list (heavy rotation playlist) tracks are played 25 times per week (presumably in primetime).
– This means an A-list track will be heard about 242 million times (equivalent to 242 million streams) per week.
– Performance royalties (PRS) for Radio 1 are apparently around £14.91 per minute, so assuming average song length of 3.5 minutes, that’s about £50 per play. This gets split 50% between the songwriters and 50% to the publisher.
– Broadcast royalties (PPL) for Radio 1 are apparently around £38 per play. This normally gets split 50% to the record label and 50% between the performers (in the case of a pop singer, most of that will go to them and a few percent goes to session musicians who also performed on the track).
– For purposes of discussion lets assume a typical pop singer/songwriter would co-write their songs with one or more people, so lets assume they have 50% of the songwriting credit (so 25% of the songwriting royalties). Also let’s assume it’s an electronic track with no other performers, except the producer takes a small amount, say 5%, for their “performance” of adding all the electronic instruments to the song (via their computer). So the total royalties this artist would receive for 242 million “streams” is (£50 x 25 plays x 25%) + (£38 x 25 x 45%).
This adds up to £740 to the artist, and total royalties on the song would be £2200 (just for that week, just for that one radio station). This works out as around £3 in royalties per million listens go to the artist, and £6 further royalties for the record label and publisher.
ARTIST INCOME = £740
LABEL & PUBLISHER INCOME = £1460
EFFECTIVE LISTENS = 242 million (non-proactive)
EQUIVALENT ROYALTY TO ARTIST= £3 per million listens
EQUIVALENT ROYALTY TO LABEL & PUBLISHER = £6 per million listens
NOTE: make sure to bear in mind that a hit song will be played for many weeks on many radio stations, perhaps around the world, so the number of actual song listens coming from radio for a hit song will be in the tens of billions. When you think of it like this, a million hits on Spotify or Youtube is nothing. Also note that big hits will get a lot of TV play, which works much in the same way as radio in terms of royalties, with extra payouts for the music video too. And there’s sync royalties for hit songs used as a soundtrack to TV shows etc, but that’s a whole different subject.
INCOME FROM STREAMING SERVICES
It seems that total revenue from Spotify (and Apple Music) is divided 30% to them, 58% as broadcast royalties and 12% as performance royalties. The 30% makes sense as it’s the same cut iTunes takes from digital sales. But it might seem odd how the royalties are split so unevenly between “broadcast” and “performance”. In radio broadcast the split is about 40:60, whereas here the split works out as 83:17. I don’t know how they decided this figure, but this means songwriters and publishers are getting less and performers and labels are getting more for streaming than for radio play.
You could speculate that the labels swayed this ratio in their favour when they collectively “invested” in Spotify, and as the labels are the master rights owners, there isn’t anything the publishers can do to stop them. This I suppose is one area of dubious morality, though note that most major labels have a publishing arm to their business as well so they’re not really losing out. In this case it’s songwriters who are losing out and performers are benefiting.
Remember, in the case of music sales, the songwriter and publisher get only 6.5%, and the label and performer get the rest. An on-demand streaming service essentially bridges the gap between radio play and music sales, so it makes sense that the ratio would be adjusted to support this fact, and on reflection I think it seems fair.
So anyway, here’s how the royalties break down:
– With Spotify as the example, over 700k streams in a week will put a track in the top 10 of the Spotify chart, so let’s assume our track hits that benchmark.
– After looking at a few different articles and reports, it seems that Spotify pays an average royalty rate of £0.0034 (0.34p) per stream.
– This adds up to about £2400 per week total royalties for our fictional hit song.
– Split 83:17 this equates to £1990 for label/performers and £410 for publisher/songwriters
– So our artist would be earning about £900 for the performance royalties and £100 for the songwriting, for a total of £1000.
ARTIST INCOME = £1000
LABEL & PUBLISHER INCOME = £1400
EFFECTIVE LISTENS = 0.5 million (semi-proactive, see note below)
EQUIVALENT ROYALTY TO ARTIST= £2000 per million listens
EQUIVALENT ROYALTY TO LABEL & PUBLISHER = £2800 per million listens
NOTE: I couldn’t find any information about percentages of songs streamed as part of curated playlists. That type of listening is similar to radio, i.e. non-proactive listening, the listener didn’t specifically choose that song. Whereas listening to an artist’s album, or even just to their singles, is proactive listening which is more similar to making a purchase. From my experience the type of people who listen to the top 40 are also the type of people who like to listen to the same songs over and over anyway, so I’m guessing a lot of spotify users like this will just play the “UK top 50” playlist on loop, or other similar playlists. To some extent this makes it a self-fulfilling prophecy that once a song gets in the chart it will stay there, and so I’d assume the percent of proactive listening on chart hits will be low, maybe 20-40%. And conversely I’d expect the percent of people deliberately listening to underground bands/artists would be quite high, maybe 80-90%. This logic suggests that streaming probably doesn’t impact chart sales too much, and probably has a bigger impact on sales of underground music. But then again, people who listen to underground music are also people I’d say are more likely to buy the music anyway to support that artist. This model would fit with most people I know, but it’s certainly an untested theory!
INCOME FROM MUSIC SOLD
A chart hit could be anywhere between 2k and 100k sales per week. On a slow week 2000 sales might be enough to break the top 40, but let’s assume the song we just talked about, after a week of Radio 1 A-list rotation, sells 20k copies (which probably puts it somewhere in the top 10).
– Most of the songs in the chart are 99p, so total revenue on our song would be about £20k for the week.
– The distributer (iTunes, amazon, etc) keeps 30% of that, leaving £14k. (Note that CD single sales are mostly zero nowadays. And also note that iTunes in the UK pays out a further 6.5% to MCPS, which will be split 50/50 between publisher and songwriters.
– Depending on the deal the artist signed with their label, apparently they will often get 8-25% of gross sales (sometimes 50% of net sales). Although from that will be deducted any advances they’ve been paid (“recoupment”).
– Also from this share their manager will normally take a cut of 10-20%.
– So this could leave the artist with something like £1200 to £7000 (just for that week, just for that one single, just in the UK, depending on all these factors and assuming there’s nothing to be recouped).
– Let’s assume an average fan will listen around 50 times to a song that they buy. (That’s in their lifetime. Check your iTunes play counts, very few songs will get over 100 listens, and there will be a lot with less than 20 listens, so I think 50 is a fair average estimate). So 20k sales x 50 listens equals 1 million listens. Giving a total effective royalty of £14k per million listens. And also note these listens will be spaced out over a few years, but these “royalties” are all paid up front in the space of 1 week.
ARTIST INCOME = £1200 to £7000
LABEL & PUBLISHER INCOME = £8300 to £13700
EFFECTIVE LISTENS = 1 million (pro-active)
EQUIVALENT ROYALTY TO ARTIST= £1200 to £7000 per million listens
EQUIVALENT ROYALTY TO LABEL & PUBLISHER = £8300 to £13700 per million listens
Note: There’s an article claiming 33 streams on Spotify equates to one sale in terms of revenue, but also the Official Chart Company have determined a rate of 100 streams equating to one sale in terms of chart performance. So again, my estimate of ‘50 listens equating to one sale’ falls in between these two, and hence I think is a fair assumption.
INCOME FROM YOUTUBE
I never really considered this before as I’m not a person who listens to a lot of music on YouTube, but after some research it all sounds very cloak and dagger… The sneaky people at YouTube (Google) have strong-armed the music industry into basically letting them do what they want, and apparently this is paying a relatively low (but undisclosed) royalty however they see fit (i.e. once every few years it seems).
The best numbers I’ve found are $30 royalty on 1 million plays, but who knows how that is split. Hopefully when YouTube’s new music service “MusicKey” is rolled out fully, the royalties will be in line with other on-demand streaming services. We can only hope.
– There is advertising revenue from YouTube hits, and this apparently can work out around $8000 (£5200) per million hits. But this isn’t a royalty, and so won’t be split according to any royalty laws but would presumably go to the master rights owner (record label)… I’d think an artist would be lucky to see any of that unless they have a very good record deal.
– If the report of $30 (£20) performance royalties on a million hits is accurate, this puts the rate a little bit above radio and well below Spotify (and Apple Music etc.)
– I couldn’t find any UK numbers, so I’ve just converted the dollar amounts for comparison. I also couldn’t find any stats on how many YouTube views are typical per week (just from UK users) for a top 10 hit song, although I’d make a guess of 500k to 1 million hits per week perhaps on average.
ARTIST INCOME = ??
LABEL & PUBLISHER INCOME = ??
EFFECTIVE LISTENS = ?? (pro-active)
EQUIVALENT ROYALTY TO ARTIST= maybe £5 to £50 per million listens, depending on a lot, maybe more if they’re getting a share of ad revenue
EQUIVALENT ROYALTY TO LABEL & PUBLISHER = maybe up to £5000 inc ad revenue
INCOME FROM ILLEGAL DOWNLOADS
Don’t forget about this one – hate it all you want, but it’s one of the other main ways people consume music!
– Estimates are that 20-30% of music downloaded is pirated. Let’s call it 25%.
– If our song sold 20k copies in a week, this suggest a further 6600 people will download the song illegally.
– And again if each illegal download represents 50 listens, this equals 330k listens.
– The income on this is of course zero to the rights holders (though maybe a bit of ad revenue to the hosts of the pirate sites!)
ARTIST INCOME = £0
LABEL & PUBLISHER INCOME = £0
EFFECTIVE LISTENS = 0.3 million (pro-active)
EQUIVALENT ROYALTY TO ARTIST= 0 per million listens
EQUIVALENT ROYALTY TO LABEL & PUBLISHER = 0 per million listens
In this specific situation for our fictional top 10 charting song, the APPROXIMATE ROYALTIES the artist would receive PER MILLION EFFECTIVE LISTENS of the song, for comparison, are as follows (in ascending order):
Illegal downloads… £0
Mainstream Radio (e.g. BBC Radio 1)… £3
YouTube… £5 to £50 at a rough estimate
Streaming service (e.g. Spotify)… £2000
Sales (e.g. iTunes)… £1200 to £7000 (depending on record deal)
When you lay out the numbers like this in comparison, the Spotify rates seem quite fair and well thought out. They pay out at a rate comparable to sales, however with the one major difference that sales “royalties” are sort of like an advance, whereas streaming royalties are paid as they happen over months or years. So in the short term it might be worse for artists but in the long term it works out about the same, and it’s certainly better than YouTube or piracy. There are a few controversial issues that Spotify and other streaming services will have to iron out, but it’s pretty clear to me and most other people that streaming is the future of music consumption so I’m sure they’ll get there with time.
Furthermore, one could presume that Spotify (backed by the major labels) would be a lot more concerned with the future of the music industry than the tech giants Google and Apple will be… remember the major labels are TINY companies compared to Google and Apple. One of the major labels, EMI, was sold a few years ago for £1.5 billion… Spotify’s value is estimated at £5.6 billion, Apple is worth £428 billion and Google is worth £287 billion. It’s clear who will win in any power struggle, and what’s good for the tech guys may or may not be good for the music guys. As a music guy, this concerns me.
So after doing the research, my verdict is that Spotify certainly isn’t evil and is probably a good way forward for the music industry, despite all the click-bait articles complaining about royalties etc. I would encourage people to use Spotify instead of the competing services Apple Music and YouTube MusicKey, thereby supporting the dwindling music industry rather than the booming tech industry.
If you enjoyed reading this, and you do indeed have a Spotify account, please do me a favour and click the link below to check out my band. We try really hard to make the sort of tracks you’ll want to listen to 50 times! Thanks!
– The number of streams to equate one sale is higher than rights-holders would like no doubt, but in another way this idea rewards artist creating complex music with longevity that requires a lot of listens, and punishes the instant fix throwaway pop songs that no-one will remember in 2 years. So that’s a good thing for art, even if it’s bad for hack songwriters.
– It seems there is a disparity between royalties paid on songs streamed by free accounts compared to premium accounts, the former being apparently a tenth of the latter. This seems obviously unfair and seems like an easy fix, the pot should surely be evenly split rather than tracking account types separately, and the royalties can be adjusted across the board. As it stands, I suppose this benefits artists who’s fans are more likely to pay for a premium service, but that seems like a weird distinction to impose on rights-holders. At least Spotify are planning to sell more ads to close this gap, so they realise it’s a problem that they need to work on.
– I have a free Spotify account on which I listen to maybe 2 or 3 albums per month, and the reason I wouldn’t currently pay for premium is because I hardly get chance to listen to new music, so I wouldn’t feel like I was getting my money’s worth. Conversely there are lots of people who have Spotify on at work all day everyday, streaming hundreds of hours of music all for one low monthly fee, maybe even to a whole office of people rather than just themselves. The artists here are losing out, getting less money per song because more songs have been played for the same subscription fee. Apparently only 25% of Spotify accounts are premium, but I would assert that anyone who listens to Spotify for more than 20 hours per month should have to pay for a premium account, they should bring back that limit that used to be in place.
– Actually to me a fairer model would be if Spotify charged £0.01 per song streamed, and you get a monthly bill, the same as with a phone bill. No adverts would be necessary, the royalties paid out would still be 70% (£0.007 per stream, double what they are now), and it would mean 1000 songs (about 60-80 hrs of music) would cost £10 to stream; I don’t think anyone could argue against that being a fair price. (Also I’d suggest there should be no royalty due if the song is skipped after less than halfway… this would hopefully encourage better songwriting in the mainstream!)
– Differences in the US music laws mean things work slightly differently for US artists. Eg radio in the US doesn’t pay any broadcast royalties, only performance royalties, though there are waves being made about this currently so it could all change. Apparently this law also carries across the pond in that US artists don’t see any broadcast royalties from UK airplay, which seems odd.
– This article is mainly talking about “top 40” type artists, normally signed to major labels. It could be further explored how this related to unsigned artists and indie acts perhaps. And one other thing that certainly becomes clear looking at these numbers… for any major artists that can part ways with their record label and start releasing music on their own, they stand to make a LOT more money per play/sale (assuming they have the status and momentum to keep getting plays without a major label marketing budget).
Thanks for reading. If you’re still here, click this:
James Grover (masses, Peak Studios, Signal Records)