The streaming debate has been a watershed moment in artist empowerment, a discussion which has allowed them discover that they can have a meaningful voice in the digital debate. Crucially it has also been a democratization of the artist voice. In the days of Napster it was only the superstar artist who got airtime to argue for (Chuck D) or against (Lars Ulrich) file sharing. Now in the days of social media the playing field has been levelled. The streaming debate has also been a coming of age for artists as business people, coming to terms with the wealth of analytics and sales data they now have at their fingertips.
Messy Data
All of this has been good and positive, and it is an evolution that I look forward to seeing continue. However there has been an unfortunate by-product of the process. With an artist posting their latest streaming versus download income data practically every week the focus has been on quantity of data not quality and, most importantly, data has often been misinterpreted and stripped of crucial context. The situation is compounded by the murkiness created by the mass of moving parts that determine how much an artist gets paid. These include: what sort of deal the artist is on, whether they are recouped, whether the artist is just a performer and/ or a songwriter, whether the label is redistributing all of its advance payments from the streaming services with artists, whether the artist is paying additional fees to distributors / aggregators and how good a deal those organizations have struck with streaming services.
The transition from the distribution era of selling units of stuff – whether that be downloads or CDs – to the on-demand consumption era is without doubt a highly disruptive process. As the overused cliché of corporate managers goes ‘change is difficult’, and the future can look even more daunting when the new world is viewed through the same eyes we understand the old one. So as much as the ‘how many track streams an artist needs to equal one download’ comparison is understandable, it is not the most useful analysis. It measures consumption models by distribution model metrics, which is as useful as comparing the speed of a car to a plane.
Setting the Right Benchmark for Assessing Streaming
This week economics academic David Touve compared downloads to radio instead – something which I did back in February. It is a useful exercise in that it places streaming in a range, somewhere between downloads and streams but it is still only part of the story. After all, radio is meant to be the discovery journey, not the consumption destination. Artists cannot afford to live off radio alone (especially not in the US).
At the other end of the scale the 1 stream = 150 to 200 downloads comparison is fraught with problems. Indeed, if you multiply the streaming average of $0.005 by 150 you get a value of $0.75 which is virtually the entire retail price of a download. Clearly this cannot be right. The retailer needs to take their cut from a download sales, then the label, then the publisher, then collection agencies, distributors, aggregators etc etc. Only a very small share of artists will ever get near that sort of rate: DIY singer songwriters. And it is no coincidence that they are the ones who have been most active in ingesting data into the streaming debate. A standard label artist can expect closer to $0.09 from a download (as publicised by Chuck D a year ago) which is only 18 times the streaming pay out rate. Though of course some artists (who are also songwriters and are on 50/50 net receipt deals, that are fully recouped) could hope to earn nearer $0.40 per download.
What is clear is that the ‘moving parts’ of individual artists’ commercial terms are so variable and so complex that they prevent meaningful comparisons between streaming and downloads.
The Consumption Analysis
So to create a more useful set of metrics to work with I have created a comparative methodology called the Consumption Analysis. This creates a like-for-like comparison between the value of a stream and a download to artists and strips out entirely the artist’s commercial terms ‘moving parts’.
This is the basis of the approach:
a) First, downloads are paid for once but played many times, so a price per listen is needed. This is determined by establishing the lifetime value of a purchased track and dividing the sale price by the total number of plays it will receive after purchase. (I have set this at a modest 12 plays per track over 3 years)
b) Next the multiple moving parts that confuse the streaming debate need stripping away to enable like-for-like comparison. Streaming and download services both pay approximately 70% of income to rights owners. We then work out how much money per stream is paid out to rights owners ($0.0112) and what the average $0.005 artist pay out is as a share of that (45%).
c) The resulting net-neutral artist-to-rights owners ratio can then be applied to downloads, and then averaged out by the total number of plays a track receives in its life time (i.e. 12 plays in 3 years)
(For sake of complete transparency you can download the entire Excel document here – please go ahead, play around with some of the key levers and feel free to post your own findings.)
Downloads and Streams are Much Closer Than The Current Debate Suggests
The net result is, working on a pure like-for-like basis, the per-play value of a download to an artist is $0.033 compared to $0.005 for streaming. Downloads are thus 5 ½ times more valuable to artists than streams. Of course this is still a disparity but it is much, much less than the 150 to 200 times value that has become common currency.
It is also worth noting that the artist streaming pay out rate ($0.005) is actually 45% of the rights owner pay out rate ($0.0112). So artists are earning nearly the same out of streaming as the labels and publishers.
None of this is meant to belittle what is a massively important issue for artists, rather it is intended to help them understand where a path to solution might lie and to have a more balanced understanding of the value of streaming. Clearly there are many complex issues that need to be addressed between artists and the numerous parties they have commercial deals with and who therefore take a cut of their digital income. There appears to be more ‘noise’ in the revenue transition from music service to artist for streaming than for downloads. It is likely that much of this is related to the fact that the streaming model is still finding its feet and that levels will balance out over time. But there is certainly a case for artists taking charge of their own destinies and getting to the bottom of their individual situations. That does not just mean with the record labels either, it means each and every platform, agency, rights body and rights holder they have commercial agreements with.
What the Consumption Analysis reveals is that on a like-for-like basis, streaming services are clearly much closer to download services than the current debate suggests. The 150 to 200 range is neither useful nor accurate. All of this should hopefully help build some more confidence in a future in which streaming plays a key role. This is even before considering the oft touted scale argument that underpins the case for streaming. Factor that in and we start getting towards a picture of an industry that could genuinely grow after years of decline, if streaming does indeed go mainstream.
I am not suggesting that streaming is the answer to all of the music industry’s ails but it can certainly bring a lot more right across the value chain than the current streaming debate suggests.
Music Industry Blog 
All very good, but misses the most important thing, which is cash flow. The main difference between a stream and a download is that when someone buys your track outright you get that payment in one lump sum almost straight away. With streaming you will have to wait a very long time for that equivalent sum to accrue.
With so many indie artists & labels living hand-to-mouth / release-to-release, relying on income from one project to fund those in the near future it’s not surprising that they’d rather have $0.70 today than (potentially, and ONLY potentially) $1.50 in 3 years time. It will be too late by then. The value of money isn’t only in the amount it’s worth, but also in when you can use it.
Cash flow is the most important thing of all for small businesses the globe over, and every one that hasn’t gone bust recognises that a smaller amount of usable money today is better than a larger but theoretical amount of money next year. There’s no reason why these rules don’t apply to indie acts.
You may have come across this wisdom yourself over the years expressed in a very basic form – “a bird in the hand is worth 2 in the bush”
The cash flow issue is an important one. One that I think both Mark and Myself have brought up before (for example, http://davidtouve.com/2011/11/22/is-a-stream-on-spotify-or-any-music-service-really-worth-less-than-an-itunes-sale/).
As you note, this matter involved the so-called “time value of money.”
That said, there is reason to ask whether the number of people who would stream a track far outnumber the number of people who would purchase a track. This is the sort of question that will only be able to address, directly, once music service grow in terms of their total population.
Dave – cash flow is certainly important but I wouldn’t go as far as to say it is ‘the most important thing’. Indeed the current point at which the debate has arrived is questioning whether streaming actually has any actual value to artists at all, not whether that value is spread over a period of time.
You do raise an important point though and I think it is great if we manage to move the discussion on from the current point to addressing more granular issues such as this. That indeed would be success.
Mark
David – the other scale related issue is whether people consume a greater amount of music on streaming services than download stores. The extant market and consumer data certainly suggests this to be true. So more music is getting consumed which means that artists should get a greater number of income ‘events’. The challenge is that until proper scale is achieved (i.e. 10’s of millions paying) it will be much harder for an artist ti feel the benefit of scale than a label simply because a label has a much bigger number of copyrighted works to be impacted by long tail scale.
Indeed. You will find in that link I dropped in the comment above that I also executed a “listen per listener” sort of comparison between downloads and streams. Phillipe Aster executed a Radio/Deezer comparison back in 2009 (in Francais): http://electronlibre.info/archives/spip.php?page=article&id_article=00525
Different people, at different spots on the planet, are converging on the objective of a common language. Whatever we eventually call the common language for comparison, I am glad to see this discussion over methods for comparison growing.
And like yourself, I really do hope all of this work surfaces these deeper discussions over “How/Why does it matter,” and “How/Why might it change,” both now and (hopefully) a few years out.
Hi Mark,
I found this very interesting. I like the way you break down artist revenue vs label and publisher revenue, and also that you are realistic about how many times a song will be played in a life time.
Last year I built my own macro calculations looking into the value of streaming, and I think you may find them useful in your continued analysis. I looked at it from a labels perspective, or an artist that is doing self distribution. The total numbers won’t change though.
http://glacialconcepts.wordpress.com/2011/10/05/spotiwhy/
Keep up the good work, I’m enjoying your blog.
-Frank
Pingback: » Rockonomic: Spotify versus iTunes, when is a stream worth as much as a sale? david touve
Frank – very interesting analysis. The household threshold is clearly an issue. Music spending in the analogue age depended upon individual purchasing not household. To drive meaningful increments of subscription spend once / if we start getting to a household saturation point there has to be internal household add ons – e.g. family plans, multiple supported devices.
All that said, we’re obviously not going to get close to that level.
Where does the author get the figure of $.005 (i.e. half a cent) for the average payment per stream to artists from streaming services? Do we actually have any reliable figures for streaming payments to artists, as distinct from labels? One of the few published figures that I know of is that of Zoe Keating, who last year reported that she got just over $.003 per stream from Spotify. But Zoe Keating is an entirely independent artist, who receives nearly all of the payment per stream (minus a small percentage – 5% I think – to a collection agency). For an artist signed to a label, the percentage received by the artist would presumably be much smaller.
David – the $0.005 figure is the average which has emerged from all of the various artist postings. Note that the amount Spotify pays out per artist is dependent upon a number of factors such as: total number of plays per month, total amount of revenue. So if the revenue / plays ratio is higher one month artists will get more money per stream. Similarly if the premium share is higher, revenue goes up, and therefore pay outs go up, assuming that plays don’t increase at the same rate in proportion. Thus when Zoe Keating was posting last year, revenue for Spotify was markedly lower than now, and the paid subs count was lower.
Finally for a good aggregation of reported numbers, and confirmation of the $0.005 average, see SpotiDJ http://www.spotidj.com/spotifyroyalties.htm
Not convinced. All the reported figures seem to be for independent artists. Can you give any examples of label-signed artists getting as much as $.005 per play? It stands to reason that if independent artists like Zoe Keating are getting $.005 (assuming it has increased from last year’s $.003) signed artists must be getting less, just as they get a smaller share of CD and download revenue. Why would the labels agree to a larger share for artists from streaming? Of course, the economic justification for the labels is that they are making the financial investment in recording and promotion, and usually an advance to the artist (to be recouped from royalties but not otherwise recoverable). Since many (most?) artists never recoup the advance, they are effectively getting a subsidy from the labels.
David – your points are all valid, which is why I’ve tried to remove the uncertainties of these moving parts in this analysis. I am comparing the two revenue streams based upon all things being even. So if an artist is only getting say 20% of the total ‘rights owner’ pay out for streaming then the same proportion should apply to downloads also. Indeed if you look at the Chuck D earnings from downloads it was about that level.
The important point here is that the difference between streams and downloads is a matter of single digit or low double digit multiples, not multiples of a hundred+.
Does that make the income from streaming suddenly high value to artists? No. But it does show that the currently small income problem is more about the current scale of the services rather than the model being inherently broken.
Well, I don’t get that from your article. When specifically discussing Chuck D, you say that an artist of this type might only get $.09 (9 cents) from a download sale, and go on to say that that is only 18 times what they would get from a single stream. But that clearly does imply an artist’s pay-per-stream of $.005, which I am questioning in the case of signed artists. In your comment you seem to agree that for signed artists the proportion of ‘rights owner’ pay would be lower than 50%, which is what you assume in getting your $.005 figure. So I am not clear where you now stand.
If artists signed to labels get the same share of ‘rights owner’ streaming revenue as they do from CDs or downloads (and why would they get more, unless the labels are being uncharacteristically charitable?), then you have to scale down the artist’s receipts by the same ratio in both cases. So if, e.g., you assume that artists get only 15% of download income (i.e. what goes to the label after iTunes or Amazon take their cut), then you should also assume that they only get 15% of streaming income.
Incidentally, I think some of the reported figures for streaming income should be taken with a pinch of salt. I note that the SpotiDJ article which you link to in an earlier comment mentions claims that the Swedish artist Jonathan Johansson earned $20,000 per month from streaming services. Does this even pass the ‘smell test’? As I pointed out in a comment somewhere else, this would require about 4 million streams a month (assuming payment to Johansson of $.005 per stream). Granted that Johansson is big in Sweden, is this credible? Looking at view counts on YouTube, which is de facto a bigger streaming service than all the others put together, Johansson is nowhere near this level of usage. Or on MySpace, where his total plays in over 3 years are under 3 million: http://www.myspace.com/jonathanojohansson So either Johansson is getting disproportionately high traffic on other streaming services, or the figures are somehow ‘misreported’. Or maybe some Swedish artists have a very sweet deal with Spotify?
The $0.005 average is exactly that, an average from the market. I have been tracking virtually every reported number myself and keeping my own tracking document. I simply pointed to the SpotiDJ site because that is available online and arrives at a similar average (outliers like Johansson aside).
As for your point about DIY artists earning more, yes totally agree which is why I wrote this in my post:
Only a very small share of artists will ever get near that sort of rate: DIY singer songwriters. And it is no coincidence that they are the ones who have been most active in ingesting data into the streaming debate.
The key point here is this: streaming services and download stores pay approximately the same amount of money to rights owners. If rights owners pays the same proportion to artists (regardless of whether that is 5%, 15% or 50%) the income to artists between the two differs only by a multiple in the region of 5 to 15 (depending on which levers you want to pull) NOT 150 to 200.
I don’t want to prolong the discussion indefinitely, but your conclusions are radical, and likely to be influential, so it is important that they should also be correct.
A problem I have with your analysis is that it is quite complex, which means there are many points where errors of fact or reasoning may creep in. I prefer to stick to concrete examples.
Suppose we take an independent, self-financed, self-producing singer-songwriter who distributes music digitally through Tunecore or some similar agent. Such an artist will typically receive about $.6 from an iTunes download priced at $.99, of which iTunes keeps about $.3, Tunecore (etc) takes a small cut, and the artist gets the rest. Of course, the $.6 is not net income, since the artist has to pay for recording costs, etc, but these are the same whether the track is downloaded or streamed. Now let us agree that such an artist gets $.005 from a stream on Spotify (etc). By simple arithmetic, the artist would need about 120 streams to equal the revenue from a download. This is not as high as the 150-200 multiple that has been claimed in the past, based on an assumption of only $.003 per stream, but it is nowhere near your figure of 5-15!
At the other extreme, consider a manufactured commercial pop star signed to a major label, with big name producers and song-writers. They will probably also have a Svengali-manager stealing most of their money, but let’s leave that out of the equation! Depending on their manager’s negotiating skills, and their perceived commercial potential, they will get an artist’s royalty (say) somewhere between 10% and 15% of the ‘cover price’ (more than that in the case of a successful established artist on a renewed contract). In the case of a download this comes to between $.1 and $.15. In the case of a stream, unfortunately, there is no clear ‘cover price’, but using your own estimates we can reach a proxy ‘price’ of about $.015 per stream (of which 70% goes to the label and 30% to the streaming service). If the artist gets between 10% and 15% of this, as with a download, their revenue per stream is between $.0015 and $.00225. This gives a download:streaming multiple of about 67: much lower than for an independent artist, but much higher than your figure. So where do we differ? I think the key difference is your assumption that all artists get around $.005 per stream, but your basis for this is not clear to me. You say it is ‘an average for the market’, but so far as I know all of the available published figures are for independent artists. Am I wrong? If so, can you give some examples for commercial artists signed to a label? If there is firm evidence, I am willing to accept it, but if you are just taking figures derived from one category of artists and applying them to a very different category, then the conclusions are doubtful.
David, I appreciate the feedback and the debate.
What I was trying to do was simplify an already complex discussion by not using scenarios of artists. But I get your point. So let’s simplify even further. If we just look at the 70% of each model that is paid to rights owners and then look at that at a per play basis, then we get to the 5.5 time multiple. So at that top level, before any monies have been shared, recouped, accounted etc etc the ratio is a decent one.
It is clear though, and this is where your examples come into play, that artists are not yet getting the same proportion of pay out as for downloads. If an artist is getting paid for digital income on a percentage of net receipts that should be consistent. If it isn’t then it suggests something isn’t being accounted for in the same way in between the streaming service paying the rights holders and the rights holders paying the artists. There are many potential explanations and indeed culprits.
It is interesting that, to my knowledge, there have not been streaming income receipts published by major label artists. Could it be that most indies and the aggregators DIY artists rely upon just don’t have great deals with Spotify et al?
So it is clear that at the moment that many artists are, for whatever reasons, not seeing as much money as they expect they should be from streaming. But at a ‘monies paid to rights owners’ level there is no reason that the disparity should be in terms of multiples of 100’s
OK, I must admit I haven’t worked through all your calculations, but even using an ‘aggregate rights holders’ approach, I don’t understand how you get such a low download:streaming multiple. I think we are agreed that rights holders collectively (whether it be labels, artists, publishers, songwriters, collection agencies, or whatever, in any permutations and proportions) get 70% of the revenue in the case of both streaming and downloads. In the case of downloads, typically priced at around $1 per track, that means they get around 70 cents per track. In the case of streaming, your own calculations suggest that rights holders typically get just over a cent per stream. So that points to a multiple of just under 70, i.e. it takes just under 70 streams to give rights holders collectively as much revenue as a download. Am I missing something?
Let’s cut to the chase … The real winners in all of this – as has been the case throughout eternity – are the aggregators – the major record labels with millions of tracks – major music publishing companies with millions of songs – Tunecore with tons of recordings and songs who get paid whether a recordings is streamed or downloaded on the Net …or not!
Little payments from a ton of sources adds up to a lot more that little payments from a few tracks… no matter how many tracks an individual artists can crank out.
Indie artists (and songwriters) are now competing with everything ever recorded – It’s all out there for consumers to chose from.
Irrespective of what anyone tries to sell you, there is very little actual community to coalesce around one off indie artists. It’s become a one to one world rather that a one to many world … everyone is making their own playlists from endless resources…
Great for consumers … not so great for artists trying to break through.
We’re now in a world where everyone is building their own personal versions of Muzak!
Well, lets just be honest and admit that when we deal with numbers “in general” that doesn’t mean we are dealing with numbers for any specific artist.
Furthermore, while Mark states radio is meant for discovery, not consumption, radio is the medium through which most music is consumed globally. Treating radio like a promo channel may ignore how people experience this channel.
In turn, services like Spotify are just that — music services. Comparisons with revenue from radio and sales are meant to be, at least for my part, just comparisons.
That said, we tend to overvalue money paid to us today and undervalue payments over the long term. Thus, my recent post on the time value of money applied to streams…
http://davidtouve.com/2012/12/12/rockonomic-spotify-versus-itunes-when-is-a-stream-worth-as-much-as-a-sale/
Reblogged this on The Many (Mis-)Adventures of Kittu Pannu and commented:
Interesting article on how streaming services compensate artists currently and what will need to be done in the future to attribute for the trends seen in the music industry.
Pingback: — Lucas Gonze's blog
Pingback: Making Dollars & Cents of Music Streaming
Pingback: Marc's Voice » 12/21/12 – the final blog in the history of everything
Pingback: And there ain’t no doubt…in my mind…That I’m gonna stomp all over you test of time….where going platinum… « Sunset Daily
I have to agree with DavidB. I do not understand how Mark is making his calculations. Even if he were right, it still does not resolve the problem of streaming depleting income from other sources. The only model I can see working for artists would be closer to $0.10 per stream, which would indeed be on par with the 12 or so times the average download will be played, according to Mark’s own estimations. But to make this happen, streaming services would either have to impose subscriptions (somewhere in the area of $300/month if I am not mistaken) or generate serious advertising revenues form outside sources. Does anyone else see a problem with this model?
Pingback: Writing or speaking about streaming music screwing artists?
About two months ago, my best friend introduced me to Palm Beach broadcasting’s 97.9 WRMF and that has been my # 1 radio station ever since. I’m always locked on 97.9 WRMF for the best variety of 80’s, 90’s and today’s music. Check them out http://www.wrmf.com.
Pingback: %post_,title% | %blog_title, %
Pingback: Music on the Move Discos - Blog Post - Music on the Move Discos
Pingback: Spotify vs Musicians: 10 things to read to better understand the debate | Firstpage Online UK News .co.uk
How does the streaming payout work? Do I have to listen to the whole track or do clicks = listens? In that case, can I just click on a song 20 times in a row and effectively pay the artist 20 times and make up for the streaming/download deficit in one fell swoop? Or is there some kind of minimum wait required between plays?
Pingback: Spotify Vs. Musicians: Who Should Triumph? - PSFK
DavidB seems to be consistently failing to divide by the number of plays. He’s only looking at the payment whereas the whole point of the analysis was to calculate a payment per play. So when he comes to a multiple of 67 he forgets to divide that by the average of 12 plays in 3 years to come to a payment per play multiple of 5.6, which is very close to Mark’s figure of 5.
Pingback: As streaming moves into the forefront; how will artists sustain themselves? | jamesosgood17
Pingback: More Spoti-fun - The Cynical Musician
Pingback: As streaming moves into the forefront; how will artists and labels eat? | Streaming View
As you learn more and more about the platform, take notes and come up with ideas on how
to use it better. Unfortunately, it’s also introduced a host of ways to violate intellectual property laws.
It’s not always easy, but if the material is for something truly important,
it’s the wisest step to take.
Pingback: Tracks versus Albums versus Stream: The time value of money and music – ROCKONOMIC
Pingback: Killing the Music Industry | Cynical Optimist
Pingback: Will You Be Able To Stream The Upcoming Snaggle Album? | Snaggle