Spotify vs iTunes - DATA FIGHT
Posted Dec 8, 13:29 with tags , , , , , , , ,

Spotify is a frequent subject people ask me about. Here are some data, and comparison to iTunes downloads.

  • In 9 months (since February), my music has half a million streams on Spotify.
  • My label Uncanny Planet earned 500 USD for this.
  • This makes me an average of 0.001 USD per track played.

Now lets look at this from another angle, iTunes:

  • My Reminiscience album is sold for USD 11 in iTunes.
  • Uncanny Planet receives approx USD 7 for each sold unit.

So then, lets zoom out and take a fascinating overview of streaming versus download, from an artist viewpoint:

  • It takes 600 album plays in Spotify to match one iTunes album download.
  • For singles, it takes 7200 plays to match a download.

Further observations

I don't know how this compares to other artists, labels or streaming solutions, but those there are my current numbers, after Spotify and my aggregator (digital distributor) Artspages has taken their cut. 

Since I run my own label, and control all my own rights, I personally end up with most of that, after splitting with vocalists, collaborators and/or sample license costs. Other artists in a typical label deal would see a percentage of that sum, usually between 10 and 50, depending on their contract.

The Spotify income is by far the largest streaming source for my music, but still only a very very small fraction of my total digital distribution income.

My expected digital income for 2009 are slightly lower than 2008, although my overall activity is rising. This is an effect of more people streaming than buying. If this trend continues, hopefully either streaming listeners or streaming rates will grow massively enough to compensate for the loss of paid downloads.

I do not have any piracy stats, but I notice my albums are continuously growing and spreading on sharing sites. I'm not too worried about piracy and file sharing, but it would be intriguing to access stats and data for this segment, especially to compare my data and web presence with data from artists utilizing other philosophies.


I am not going to conclude anything, these are just observations; raw numbers and facts. It will be interesting to compare these rates at various points in time.

I expected the road to the future Shangri-La of digital music to be a long and windy one, and it certainly is. We are climbing along treacherous ravines, and speeding through narrow mountain passes. But it is a very exciting road, there are magnificent views up here, fresh air, and even if there is no secret green paradise at the end of the road it certainly is a great trip!


#1, by vonux on December 8th 2009, at 22:21
I don't know if I'm an average Spotify user or not, but in the last 30 days I've listened to approximately 1300 tracks (or 43 per day.) One month of premium subscription costs ~ £10, which leaves ~ 1p per track played.

At my listening rate that means you receive less than one percent of my money, which is obviously not fair in the long run.

I think in the end Spotify just need to get more users to pay for the service. The advertisement version is good for recruiting users, but at some point the users need to be converted into customers. For me the S60 client made the difference.

#2, by GMM on December 8th 2009, at 22:42
Good angle, seeing it from the listener perspective, thanks for writing that.

I have talked to a few friends regarding their Spotify habits, trying to find out how people use it. I was happily surprised to learn that quite a few are actually paying the premium service. But nevertheless, it is really hard to find some kind of "average user data" that represents how much you get/use/play for what you pay.

Myself I have the premium, and I listen to tracks way less than you. Maybe a few times a month it runs for a whole evening, other than that it is sporadic use, a few tracks here and there. If I have to make up a number, I'd say I play on average below 20 tracks a day, half of yours.

#3, by Juuso on December 8th 2009, at 22:58
Excellent post Gisle. Unfortunately the way I see how Spotify "works" for the artist is that it doesn't work. The math just doesn't add up, no matter how you spin it.

Even if EVERYONE had the paid subscription, artists still wouldn't make money. That is, after you do the real world math. Like said, average user maybe streams 1000 times per month which leaves about 1 cent per one stream to the artist. Which ofcourse is ridiculous compared f.ex to iTunes.

I'd recommend for everyone to read this article;

#4, by vonux on December 9th 2009, at 08:39
@Gisle That means your money will go twice as far when it comes to pay artists, which is good :)

@Juuso The question is how much money would Spotify be able to pass on if everyone had the paid subscription. If they could pass on 50%, Gisle could already make a (very) basic living off of Spotify alone.

Now 50% might not be realistic, but is it realistic for anyone but the major names out there to be able to make a living from any single revenue stream out of the many there are?

Finally: Every user out there have different habits, but personally I now spend more money on music than I did before.

I also listen to a much higher number of artists than I did before, which means Gisle might from now on see less of my money than he did before, while the-industry-as-a-whole sees more of my money than it did before.

Less per listener * more listeners = get the factors right and it works, get them wrong and it doesn't. I just hope the experiment will be allowed to run long enough to actually figure it out :)

#5, by Khake on December 9th 2009, at 10:25
Regarding Spotify:

The app itself is total WIN - streaming content with a LARGE pool of music - now also available for offline use with various mobile apps. TOTAL WIN.

However, I think they have got their strategy slightly wrong - as also indicated by vonux;

The free version obviously does not generate enough income for it to be a viable solution in the long run. I notice that the number of commercials that generate money vs commercials for Spotify ("This is Rebecca from Spotify") that I suppose only generate potential new customers are disproportionate. I.e, there are not enough "external" ads to generate income (also income for the artists)

Now, the free version is really cool and a good way to demonstrate the product. But I believe that the free version should come with a limitation on how long you can use a free version of Spotify, say 3 months or something like that. After that, either use the Premium version or stop using Spotify. This is obviously numbers taken from the top of my head, but something like that. You get the general idea, at least.

At least to me, it seems obvioous that for Spotify to make money, and consequently pay musicians and artist, they are dependent on PAYING users.

That said, I only use the free version of Spotify - so I am aware of the irony of what I am saying. Having said that, I personally BUY albums in a physical format - either cd's or vinyl. I use Spotify every now and then to either check out new music and/or when I'm at parties where no music (or no interesting music) is available.

And regarding the piracy issue - which has been debated too much anyway - I am in principle opposed to it, but still download pirated music myself. But most of the time I do this either to check out an album I consider buying, or while waiting for the actual release on my preferred physical format; vinyl. As is the case with for instance the fantastic new release Nekropsalms from Norwegian death metallers Obliteration. Released on cd in November, vinyl release in January or February. So until the vinyl is released, I'll settle for mp3. I actually don't see anything wrong with that.

My 2 cents.

#6, by PiNaCZ on December 9th 2009, at 14:13
Hi Gisle,

since I am a member of the tracker I can provide you with some "stats" for Ugress and Nebular Spool albums there.

Each line combines download of all formats (V0, V1, V2, FLAC) that are available for particular album.

I dont condone piracy but it is a nice place to find new/similar artists and to be able to try before I buy.

Greetings from Prague,


#7, by GMM on December 9th 2009, at 18:00
Thanks Martin, great with numbers!

It's a bit hard to put them into scope - I am not familiar with, I suppose it's a torrent tracker of some kind... are those numbers representative for general torrent piracy?

#8, by Praxis on December 9th 2009, at 20:54
Interesting about Spotify!
Most of us is quite clear about the "fact" that S. does not generate enough for the artist at the moment. I guess there will be more paying customers, more ways of buying, and more paid spots.
I work for a company that will be spending money at Spotyfy in 2010. So far the service have been lacking targeting options, but things are getting better.
I think artist should be a bit patient with Spotify, because they are breaking new ground for music listening. But at the end of the day, the service need to generate more for artists, labels and so on.

#9, by PiNaCZ on December 10th 2009, at 09:58 is currently the largest music torrent tracker (private), check out its stats from today:

Actively Browsing: 1,680
Maximum Users: 125,000
Enabled Users: 109,676 [Details]
Users active today: 33,488 (30.53%)
Users active this week: 69,614 (63.47%)
Users active this month: 94,712 (86.36%)
Torrents: 664,541
Albums: 327,403
Artists: 254,146
Requests: 71,349 (63.85% filled)
Snatches: 25,165,062
Peers: 3,779,853
Seeders: 3,684,200
Leechers: 95,653
Seeder/Leecher Ratio: 38.51

Most of its members are music lovers so it should give you quite good insight of how popular is your music among them. From my point of view it is good to be there to be discovered ;-)


#10, by opyrt on December 19th 2009, at 22:55
I too hope to see Spotify being able to pay artists what they deserve, but I really don't like the combarison of "number of streams on spotify" vs "a buy on any other platform".

Say I'm totally addicted to an Ugress track and listen to it 500.000 times on Spotify. Or I buy that track on itunes. Should the first alternative pay thousands more to the artist than the other? A number I have yet to see in this kind of discussions is "completed unique streams", which is much more relevant.

GMM: If Spotify gives you those numbers, I would love to see how they compare to the number of streams.

Keep up the good work! :)


#11, by frozenintime on December 25th 2009, at 02:20
heya GMM - interesting thread!

ive been known to share your music on bittorrent sites over the years - but only @ 96kbit sample.

that way, people who hear of you, and seek your music can dload it and if they like it enough, they can buy it in full quality - just like I did years ago, even before you had your website. I remember dial up dloading on limewire because i had heard one of your tracks, then purchasing your CDs (which i still have and treasure) - the postage cost me more to australia then the CDs!

Merry Holiday to you - thankyou for the music :)

#12, by Asbjørn Ulsberg on January 4th 2010, at 11:28
Very interesting article. One dimension I miss in this comparison, though (and all others I've seen; it's always download versus streaming), is radio. I'd say that streaming (Spotify) lies somewhere between radio and download (iTunes) in terms of user experience. What would be interesting to know is how it compares in terms of income.

Does streaming yield more income per listening audience than a typical play on the radio? It's of course hard to measure given the relative extreme difference in number of listeners a song can have on one radio show compared to another, but say prime time playing in the number one radio channel in Norway -- how much money would you get?

#13, by GMM on January 4th 2010, at 12:25
Interesting question, Asbjørn. You can see the radio stats here (Norwegian):

The first document contains the stats for 2009. My music is usually in group 1, so:

Played on NRK P1-P3 during daytime; NOK 73 pr min

It is hard to find concrete stats for number-of-listeners for NRK, but I found this, stats for summer of 2008:

Where it looks like there are on average almost 2 million listeners on NRK P1, and 300 000 listens to NRK P3. Since P3 is most likely to play my music, let's use them, this gives the following income-per-listener for radio:

73 NOK pr minute
A song runs for 3 minutes, totals NOK 219
300 000 listeners
Average income is 0.00073 NOK pr listener pr song played (0.00001 USD)

I'm sure these data are sort of questionable, I didn't research this deeply, but this little experiment tells me that I POTENTIALLY make 10 times as much per person with Spotify, than with national radio. This, of course, does not take into consideration if I am played on radio or not. It's just an average number for the discussion, not a result of my personal income. But it does confirm my belief that direct access and streaming and personal reach is more important than hits-on-a-radio-channel (but you kind of need the radio to REACH those individuals first, so it's a catch 22...)

#14, by Asbjørn Ulsberg on March 23th 2010, at 08:48
Those numbers are very interesting and while they aren't highly scientific, they do confirm my suspicions and make it kind of obvious that streaming is closer to radio than download, not only in the user experience, but also in in income. At least for now.

Thank you so much for digging into the numbers and for serving the world so relentlessly with your truly amazing music! Keep up the awesome work!

#15, by Yngve on August 5th 2010, at 18:37

Very interesting numbers. And with that I start to question more why the labels seem to be so disappointed in Spotify. They are still having a hard time getting "normal commercials" instead of the "Hello this is Rebecca" ads, and are already potentially beating radio by a factor 10.

You are sort of right about the catch 22, but Spotify could potentially give you the same reach. E.g. by music sites on the web creating playlists with your song in it, which then many people subscribe to. I already subscribe to many playlists with thousands of subscribers. I could easily imagine someone like e.g. NRK creating a playlist with "Ti i Skuddet" which could have tens or hundreds of thousands of subscribers. You then have the same reach as the radio... It is just thought up idea from my side at the moment, but I think the potential is obviously there.
I think labels have (intentionally?) misunderstood some of the piracy concept from the get go. They have almost exclusively considered one download to equal the loss of one CD sold. What is correct however is that many have continued to buy music (me included, actually my interest in music increased tenfold after I started downloading music), and used piracy as a way to check out new artists before buying.
Now I have Spotify, which has killed my need to pirate, but not my "need" to get the good CDs in my collection :)

#16, by Malcolm on September 15th 2010, at 14:39
This is great news. Your earlier posts about Spotify led me to believe there was no financial benefit to streaming audio services, but it looks like things might be turning around now.

I wonder what your Spotify income would look like if the service was available outside Europe?



#17, by Yngve on October 26th 2010, at 21:56

Excellent statistics, and as you say there are no indictations in your data that the iTunes sale is suffering from the increased income on Spotify. Are you considering to get your music on other streaming services that are available in the US (or are you already perhaps?) Rdio is sort of similar to the model of Wimp from what I've heard?