Growth of Paid Downloads vs. Streaming, 2012 vs. 2011…

Cannibalism, defined (in capitalism): the process by which one closely-related, oftentimes evolved product reduces demand for another.  The opposite of complementary.

Growth of Paid Downloads vs. Streaming, 2012 vs. 2011

A. Paid Downloads: 1.336 billion. Up 5%.

(iTunes, Amazon, etc.)

B. Spotify Subscribers: 5.2 million. Up 86%.

C. Spotify Active Users: 20 million. Up 100%.

D. Deezer Subscribers: 3 million. Up 114%

E. Sirius XM Radio Subscribers: 23.4 million. Up 6.9%.

F. Pandora Listening Hours (per qtr): 3.56 billion. Up 67.9%.

Paid download data according to Nielsen Soundscan, US-based, all other figures based on public company disclosures.  Spotify current subscriber figure estimated based on straightline growth trajectory.  And, here’s a complete breakdown of subscription subscriber levels.

Written while listening to Mozart’s “Sinfonia concertante for Violin, Viola and Orchestra in E Flat, K. 364.”

28 Responses

  1. Visitor

    So, basically the problem is the existence of smartphones/always-connected devices.

    • IFPI

      The music industry really hope that subscription streaming take over

      50 million paying subscribers at $120 a year = $6 billion

      and the royalties rate is 70% so $6 bil x 70% = $4.2 billion


      On average each person spend about $23 a year on recorded music (USA stats).

      Would you rather have $120 or $23?

      • Faza (TCM)

        I’d love to know where you intend to find 50 million paying subscribers at $120 p.a.

        As things stand, Spotify only manages to get approx. 20% of its users to pay (at both tiers) – meaning something like 600,000 paying users. Pandora One subscribers, on the other hand constitute a single figure percentage (I estimate 2%, give or take – source of calculations here: of its total user base. I don’t have the figures for everyone else, but I’d be rather surprised if they were very different.

        To put that in perspective, the average U.S. Spotify user is actually worth only around $18 in music spending (Calculation: average of $120 and 60$ tiers – that really should be weighted, but I don’t have a breakdown – times 600,000 paying subscribers, divided by 3,000,000 total subscribers), while the average Pandora user is only worth $0.77 in spending (Calculation: 1 million PanOne subscribers times $36 p.a., divided by 47 million active Pandora users).

        And here we come to the big issue with average spending: it averages out. If you have an average spend of $23 p.a. across the entire U.S. population (that’s 315 million people, or so) – it doesn’t mean that everyone spent $23 dollars. Instead, most people spent a lot less, while a minority spent a lot more.

        I find it incredibly difficult to believe that people will sign up for subscription services in order to spend more money on music than they are spending right now – that would require assuming they are stupid. Instead, I am inclined to believe that heavy music spenders are signing up for streaming in order to spend less money – especially since the vast majority of streaming users pay zero, zip, zilch, nada.

        • Visitor

          Really, it is stupid to pay more for a better service? $120 a year gets you a year worth of unlimited ad-free access to 30 million songs tracks on a ton of devices… $23 of downloads gets 23 tracks.

          • Faza (TCM)

            Perhaps I should be more explicit: paying $120 for unlimited access to 23 tracks is stupid, if you could get the same unlimited access by buying downloads for a total of $23 and I can’t imagine anyone doing that.

            The $23 dollar figure is an average anyway, so we’d be hard pressed to find anyone who spends exactly that amount annually. Instead, we have a lot of people spending a lot less and a small minority spending a lot more (the Brit term is the Fifty-Quid Bloke – an moderately affluent individual who spends around £50 a week on entertainment; as you can easily calculate, that works out at £2,600 or around $3,900 spent on entertainment annually).

            It’s folks like the Fifty-Quid bloke who are most likely to sign up for Spotify, since the £120 (UK top-tier) gives him access to something like 17 million tracks, while previously it would have bought maybe one album per month (bear in mind that this doesn’t imply that the FQB was previously buying merely 12 albums a year – he might well have been purchasing double or triple the amount). The tenner he spends monthly on the service only works out as 5% of his entertainment budget, which is a bonus.

            On the other hand, the college kid that gets all his music off the Pirate Bay or some such isn’t suddenly going to start spending $120 to get a smaller selection of music than he’s already getting for free.

            The bottom line is that Spotify or any other such service isn’t suddenly going to make the mythical ‘average consumer’ start spending $120 a year on music. Realistically, it’s far more likely that it will result in average consumer spending on music decreasing (because those who used to spend more – like the FQB – are spending less). And that’s even before we get into how that spending is divided between artists, which is an even smellier kettle of fish (and no, this isn’t the result of Spotify or the labels pulling a fast one – simply that the model is horribly broken).

          • Visitor

            This is a silly arguement sir. Maybe the rich bastard would pay $200 for your album instead of $10, so how about you charge $200 for it in case? That will surely work out.

            You can’t charge the maximum people are willing to pay, you can only charge an amount that will maximize your return. And unless you play the kind of music rich people like to hear, I doubt the amount you can charge is high. I’m pretty sure the world isn’t filled with people that spend 1/6 the average income on entertainment. Just a thought.

          • Faza (TCM)

            I’m sorry, but it is your argument that is silly and you also don’t seem to be paying attention.

            Let’s start by saying that music (and entertainment in general) is a luxury good in any case. In economic terms, this means that any spending in this area comes out of discretionary income, i.e. that portion of a person’s income that is left over after they have paid their cost of living (plus any additional essential costs and savings – for example, putting money away for their children’s tuition fees).

            What follows from this is that an entertainer is primarily targetting those with at least a moderate discretionary income – a subset of the population. The $23 average spending figure, on the other hand – and I cannot emphasise this enough – is the spending of that subset divided by the entire population of the country. It seems likely, to me at least, that the majority of the population spend less than $23, while a minority spends a lot more – driving the average up. Think of it this way: if someone buys ten albums off of iTunes over the course of the year, he is making up for three people who’ve spent $0 on music over that same year: (9.99×10 + 0x3)/4=24.975. A single top-tier Spotify subscription is enough to cover four people that spent nil. In short, if we’re considering people prepared to pay around $120 for music annually, then a mere 20% of the population can generate enough spending for us to get that average.

            The key thing is that any lower-cost alternative will only decrease spending – never increase it. The only way to actually increase spending accross the population is to introduce a new offering at a price point that is accessible to more people, but at the same time sufficiently inferior to keep higher spenders from switching to the lower-cost offer. Otherwise any gains in the lower-spending part of the population will be offset by losses in spending from the more affluent portion.

            We know this to be the case by seeing what happened to music spending over the last decade. In 2000, average recorded music spending across the U.S. population was ~$51 (value data: RIAA 2008 Consumer Profile, population data: U.S. Census Bureau – Population Distribution and Change: 2000 to 2010). In 2010, it was ~$22.5 (value data RIAA 2011 Year-End Shipment Statistics, population: as above) – it has essentially halved. While this has happened, numerous lower priced alternatives were introduced: file-bartering piracy (obviously), iTunes – with its pricing structure that not only lowered the unit cost of albums, but actually encourages pick-and-mix singles purchases – YouTube (which works fine as a free, on-demand music streaming service), internet radio, streaming services and the like. You can get the same amount of music (or more), for a lot less money.

            Aside: The Fifty-Quid Bloke need not be a “rich bastard” by anyone’s definition. Try more like “moderately young, middle-class person with no kids”. Plus, the fifty pounds a week goes towards all entertainment – not just music. In the end, the FQB might only be purchasing 2 albums a month (call it £20 – just 10% of his total entertainment spending) and still spend twice the amount as he would have on top-tier Spotify. To say nothing of the fact that his album purchases might mean something like a pound each to 24 artists, while his spotify spending might mean something like a couple of pence each, payable to a hundred artists. Spotify simply isn’t a viable income source and never will be.

          • wallow-T

            For once I mostly agree with Faza. However, what his discussion of the 50-quid bloke omits is:

            a) The rise of lots and lots of other entertainment clamoring for some of that 50 quid, primarily home video and games, and most recently the premium 3D movie ticket.

            b) The ongoing decline in middle-class living standards in this century. Gasoline is never going back to $1.50 per gallon. Taxes this month take an extra 2% bite off every American paycheck (end of Social Security tax holiday from 2009). More and more health care costs are being piled onto the employee; my family health care spend has gone from near-zero to about $200/month.

            THIS American “50-quid bloke” found, starting in 2008’s gas crisis, that he had no choice but to slash the 3-album-a-week lifestyle he had enjoyed for over 20 years.

  2. Visitor

    But here’s the good news:

    We created the monster, and we can destroy it.

    Without you, there’s no Spotify.

    • Visitor

      Without the major labels, there is no Spotify. Without you, there is simply one less artist.

      • Visitor

        Without artists, there are no labels. The opposite is not true anymore.

        Today’s artists have more power today than ever before — and they know it.

        • Visitor

          I’d rethink that…

          If the Internet is working for Musicians, Why aren’t more Musicians Working Professionally?

          • CHoc Donut

            THere’s nothing to rethink. Artists have no choice but to keep their work out of the hands of ripp-off services like streaming services. The only choice is the humiliation of getting paid 25 cents for ten thousand plays.

            There’s still plenty of internet radio that is popular as well. Streaming services need to be castrated. This is not just a battle about music, its about all copyrighted material.

          • CHoc Donut

            THere’s nothing to rethink. Artists have no choice but to keep their work out of the hands of rip-off streaming services. The only alternative is the humiliation of getting paid 25 cents for ten thousand plays. I’ve dealth with some of this on the photo side.These clowns want to pay you .25 cents and act like your friend. My response to that is a hearty F yourself. You come into work and make 10 cents an hour, you dhead.

            There’s still plenty of internet radio that is popular as well. Streaming services need to be castrated. This is not just a battle about music, its about all copyrighted material.

  3. M

    “The problem”?? What are you talking about? Streaming is the solution to the piracy problem. This is the future. Songwriters and artists are actually getting paid when you use Spotify.

    What’s the alternative? Forbidding the Internet? Forcing people to buy music on a medium they do not want?

    • Visitor

      “Forcing people to buy music on a medium they do not want?”

      Um, how do we do that?

      I see no guns… no threats, no violence.

      Please explain what you mean by ‘forcing’. Try to be as specific as you can.

  4. hippydog


    Wheres the Cannibalism?

    the “smoking gun” per se?

    Yes.. Digital downloads have slowed down..

    CD sales have dropped significantly year after year..

    But the numbers dont match..

    When you include some leeway for the economy, Ageing demographics (baby boomers dont ‘buy’ much “new” music), cannibalation from other sources (money spent on movies, video games, apps for phones), etc etc etc

    It doesnt look like there is hardly ANY cannibalism..

    Im not saying there is zero effect.. it just doesnt seem ‘as bad’ as you might expect..

    We probably wont be able to see any ‘real’ trend until 2014-15 where maybe things will ‘normalize’ a bit..

    • Visitor

      “Wheres the Cannibalism?”

      You sure you saw the pretty graph? 🙂

      • Casey

        The graph shows growth. Despite the big push for streaming, download sales still increased. If cannibalism is happening, it is not very significant.

        • Faza (TCM)

          Pro-tip: Cannibalisation doesn’t have to manifest itself as an actual decrease – it’s perfectly enough if paid download growth is smaller than it would have been in the absence of the streaming alternative.

          How do we tell? We can compare growth the growth rate of downloads prior to the rise of streaming alternatives (e.g. before and after the introduction of Spotify to the U.S.) and see whether the trend has changed significantly.

          (Answer based on quick research: Maybe it has. While there has been a significant slow-down in download sales since 2008, growth year-on-year single-track sale growth in 2009 and 2011 has been over 8%. So last year, we’ve seen just over half that. However, the snag is 2010 where year-on-year growth was just 1%. Without additional data – such as pertaining to that year’s releases compared to others in the chain – we can’t really say whether it was an outlier or if it’s just normal fluctuations. Album sale figures may provide additional insight when they become available.)

  5. Me

    Why are you comparing the number of downloads to the number of subscribers? These are two completely different statistics.

    • paul

      I’m really using whatever I can get my hands on. Pandora reports total listening hours, Spotify has subscribers and active users, Sirius subscribers, etc. So, to the extent you want to say ‘apples and oranges,’ those are the fruits I can buy.

      Beyond that, there may be stuff floating out there, please share if so (everyone…)

      Also, worth noting that YouTube may reveal updated, 2013 numbers as well which will be interesting.


  6. Matt Pincus

    Couldn’t this just be the relative maturity of the different business models? Not hard to grow 80% + from almost nothing. Much harder to grow 5% from billions of transactions.

  7. JKalinowitz

    Wish YouTube + Vevo were included, but still worth a look

  8. wallow-T

    An anecdote: A 60-something friend told me about her 30-something son. She wanted to get him some specific music that she thought he would like. She is a tech early-adopter, she was into iTunes and iPods years before me; it’s not like she wanted to give him a CD.

    For her son, however, even owning and caring for iTunes or Amazon music download files is too much trouble. He does not want downloads. It’s streaming or nothing for him. Eventually she caved and got him a Spotify subscription.

    Just one anecdote, but: reminiscent to me of the story circa 2002 when a major label invited some teens into the home office for market research. On the way out, the teens were invited to help themselves to promo CDs, but almost no CDs were taken.