Spotify Is Broken: The Lie Of ‘Feels Like Free’

One of the big questions hanging over Spotify for me has been ‘do premium plays pay more than Spotify Lite plays?’ – I.E., do I get paid more if someone with a premium account plays my tunes vs. someone using the ad-funded version.

It stands to reason that the person with the premium account is paying more to listen, so surely you’d imagine that’d be reflected in the royalties?

At SXSW this year, the CEO of Spotify was giving a talk, I asked the question about royalty rates via Hugh Garry and apparently they are distributed evenly.

This is, as far as I can see, Spotify’s MASSIVE mistake. A deal-breaking, game-not-changing, screw-up of gargantuan proportions.

Here’s why.

The people best placed to promote Spotify are artists. We can link to it from our sites, we can provide links to it when we release new music, we can blog about how great it is and share music by our peers via the links.

If we push it, it becomes the place to find our music.

Spotify needs premium accounts for it to work. At the moment, their strategy for getting people signed up is to annoy the shit out of you with adverts until you capitulate. So you get irrelevant adverts that provide no value at all to the user, and therefor no value to the advertiser. Ergo, the amount paid per advert is likely to go down not up, killing the ad funded model. If I was an advertiser there’s no way I’d bother with Spotify. ‘Can you pay to produce an advert that we’re going to use to annoy people into paying not to hear it?’ no thanks.

So what would work? Spotify’s (and the other streaming services) best chance of success is if artists see it as a viable alternative to selling individual albums and tracks digitally. If it becomes that, the amount of traffic will go up and all that listening will be happening in a discovery environment, so more music will be heard by more people.

They could also make way more if the ads were something other than anti-value annoyances to be got rid of. There are loads of ways of making ads work in this setting – referrals, targeting, favouriting, user-profiles, profit-share, in-browser special offers… all kinds of stuff that would make the ad-side of the site self-supporting. If it isn’t currently viable, then the solution is to up the level of the ads even further til it is viable. The listener needs to FEEL what their listening is actually costing.

Why? Well, contrary to what Gerd Leonard has been telling us for years, ‘Feels Like Free’ is not the answer. It never has been and never will be. Free is, in fact, better than ‘feels like free’. I’d rather make my music free to download, no strings, and be rewarded in gratitude than have some weird filtered, taxation-based payment mechanism for it where people are left thinking music has neither cost nor value because there’s no tiered pricing, no opportunity to ‘pay what you like’, no thought about the value over and above the experience that access is via a portal and detached from the artist…

Listening to ads is a form of payment. We all know that. If the ads don’t cover it, then it’s a lie to keep that system going by subsidising those listens from people who are actually paying – people who are quite explicitly paying a subscription rate that puts a distinct value on their listening time. To not divide those up is to say that the value of both listens is the same. It isn’t.

  • Spotify Lite is a limited but hugely useful discovery platform. If you have the kind of life where Spotify Lite is ‘enough’, then you weren’t about to pay £10 an album for CDs anyway. You’re probably the kind of person who listens to the radio and buys the occasional compilation. Certainly not the kind of person for whom £120 a year for Spotify premium is workable.
  • Spotify Premium is an alternative to buying music. It’s also, when you look at how long people spend listening to music, a great model for paying a sensible amount per listen. If – and only if – it’s not being used to prop up a broken ad-funded ‘feels like free’ bullshit model.

If you want me to pay £10 a month for music, let me allocate where that £10 goes by choosing what I listen to. Make that £10 count, make it mean something. Cos otherwise, I’m going to stick with eMusic, where I know that my monthly sub goes to the people whose music I’m downloading. I know they get a set amount per track, that they wouldn’t get if I wasn’t paying for it. Real end to end value.

‘Til then, there’s no way on earth I’ll be paying for Spotify premium, and I won’t be encouraging anyone else to either.

If this feels like a deal-breaker to you, and you already have a premium account, you might want to consider cancelling it, and emailing Spotify to tell them why. Or better yet, blogging about why. Let’s have this discussion in public where possible.

[and before the inevitable ‘hey, I thought you loved Spotify!’ comments happen – I still think Spotify-lite is an awesome discovery tool. Spotify premium is, as yet, way too small a slice of anything to make me rethink my position on that. I don’t need to make money from Spotify-lite for its value to be realised. But the payment model that’s there doesn’t work, so the growth curve that Spotify needs to remain viable will be a seriously uphill struggle.]

49 Replies to “Spotify Is Broken: The Lie Of ‘Feels Like Free’”

  1. Very interesting post steve,thanks for it.

    But I wonder – you say you don’t recommend paying for spotify at all. But i’m someone who hasn’t paid for music at all in 5+ years that recently signed up for there service and am happy to pay. Should i stop and go back to my old ways? Would that help artists and music? Would like to hear your advice on this.

    1. Hi Rick,

      I don’t say you shouldn’t, I say I can’t see any reason to, or any reason to recommend it to other people. But this is a grown up place, and lots of people here disagree with me 🙂

      I think regardless of whether you pay for things or not, I think you should make friends with – or at very least get to know more about the lives of – the people who make the music you love. In your relationship with them, an answer to all this will become clear.

  2. Hi Steve,

    I enjoyed reading your post, and I’ve been thinking about it for a while. I want to ask you about the premise for your argument, that

    “the person with the premium account is paying more to listen, so surely you’d imagine that’d be reflected in the royalties?”

    It is true that the person with the account is paying more, but it’s not necessarily true that Spotify is receiving more revenue for that stream. The ad revenue for a Spotify Lite stream could equal the subscription revenue, so it doesn’t follow that the increase in price to the end-user is reflected in the royalties. Now I’m sure that currently the ad-revenue won’t equal the subscriber revenue, but it’s probably quite close. (I work at, where we’re having some success at matching the costs with the revenue from advertisers.) I know it feels like the end-user’s money is worth more than the advertisers’, but unfortunately that’s not the case 🙂

    One other thing: I am unsure what is meant by Spotify’s ad-supported and subscription-supported royalty rates being “distributed evenly”. I suspect it doesn’t mean that they are using exactly the same cost model. I can only speak for we7, but our contracts require us to keep the subscription revenue ring-fenced. It is shared out on a pro-rated (by number of streams) basis to the labels: i.e. if label X’s recordings account for 21% of subscribers’ streams, then label X receives 21% of (an agreed share of) the revenue. When we work that out as a per-stream value it *is* more than the ad-supported rate, but not by much – certainly nothing like the difference between free and a tenner a month!



    1. Hi Mick,

      thanks a lot for your comment.

      The ‘even distribution’ point was in response to a question that was asked about royalty rates comparing Lite and Premium at SXSW – the report came back that Daniel said all plays paid the same rate. If they don’t (either because he misunderstood the question, or the term ‘royalties’ doesn’t cover all the payments that are made under each kind of account) that changes things – by how much depends on the accounting. I appreciate you being open about We7’s accounting practice…

      As for Subscriber vs ads, are you suggesting that when divided up, the ad revenue for *one* play (ie, the money earned from that listener listening to that ad) is the same as the fraction of a tenner (minus hosting/streaming/admin/% profit to We7)? Or do you mean that the total of all ad revenue equals the total of all subscriber revenue? Because in order to get that Ad revenue, I’m assuming you need WAY more ‘ad supported listeners’ than the number of premium account holders you need to make up that figure… or is each ‘free’ listener REALLY worth a tenner a month to you???

      IE which of the following equations results in the biggest number:

      number of ad-supported listens ÷ total ad revenue = ???

      number of premium account listens ÷ total premium revenue = ???

      Here’s my problem with all of this – and, by extension the idea of flat licenses put forward by some people (HUGE post on that coming soon) – it breaks the link between value and payment. It strikes me as a very utilitarian arrangement: musicians need to make *something*, but it doesn’t matter how they get it… If even putting up with ads doesn’t cover the money required for me to listen, then the listener is in negative equity – their listening experience is falsely defined, because someone else who is paying for the music (for whatever reasons, however altruistic or self-focussed) is making it possible. I don’t really want to be a part of that. I don’t want someone else paying for my listening. If I subsidise someone else’s listening, I’d MUCH rather do it by buying them some specific music, by paying for them to hear music that I want to share with them…

      The absence of digital facilitation of a mapped relationship between social value and monetary expression is problematic to me… it’s possible that I just need to suck it up and get used to it, but the fact that it’s REALLY not what I want as a fan, moreso than as an artist is what troubles me…

      I need to think about it some more. Thanks again for your comment, I really appreciate your thoughts and information!

      1. “As for Subscriber vs ads, are you suggesting that when divided up, the ad revenue for *one* play (ie, the money earned from that listener listening to that ad) is the same as the fraction of a tenner (minus hosting/streaming/admin/% profit to We7)?”

        Yes, pretty much – those numbers are comparable. BTW I was once a physicist – I have a habit of considering numbers comparable if they’re within a factor of 10 🙂

        When it comes to your two equations the latter does result in the bigger number, but not by as much as you might assume. And I’ll add a caveat to all of this: we’re in the very early stages with subscription at we7, and so these early indications may not be representative.

        Your next couple of paragraphs cut to the heart of what we’re trying to do: match up the values and payments for all of users, advertisers and artists (via labels and PRS etc) in a way that keeps everyone happy. You’re thinking hard about the balance between the artist and the user. Consider for a moment the balance between the advertiser and the user. He’s paying for something that the user often values at or below zero… (Of course those guys don’t have souls, and so don’t matter 🙂 )

        You’re certainly not the only one to be uneasy about it, and I don’t even begin to know the answers. But here’s the way I see it – these new services are an experiment to see what can be made to work. Naturally, not every turn will be a good one.

        “I don’t want someone else paying for my listening.”

        Then you have a big problem with ad-supported music in its own right 🙂

        Hmm… there’s lots more to for me to think about.



      2. “I don’t want someone else paying for my listening.”

        Then you have a big problem with ad-supported music in its own right

        …indeed I do, though given the multicurrency way things work online (cash is only one way of expressing value, and perhaps not the most useful or important one in many situations) I see the willingness to sit through an ad as being a payment.

        …that’s from the listeners point of view. The value to the advertiser from me is pretty much zero. There isn’t a single thing I’ve heard on Spotify that I’ve in anyway interacted with. Many of them I’ve laughed at and ended up feeling more contempt for than when I started… In that respect, I’m perhaps unrepresentative, but you have made me think about whether I should have my music on an ad-supported service at all… not out some high minded boycott, but more just as an expression of the world I want to see, not just the one that’s offered to me…

        Thanks for (inadvertently?) taking the conversation in a more spiritual direction. 🙂

      3. What about if Spotify (or other ad-funded music services) started getting the targetted adverts thing more together? For example, perhaps adverts could have a “like” and “dislike” button. That way you could mute ones you particularly dislike (a win for you), the advertiser at least gets some feedback (a win for them) and Spotify get data which will enable them to make better guesses about what fits where and and also a tool to measure that (a win for them and probably a PhD thesis for someone in information science, too!).

        Another possibility would be if you could create an ad-list alongside your playlist and pick the adverts you most want to listen to. Not everyone would use it but, again, it could be a three way win.

  3. “What about if Spotify (or other ad-funded music services) started getting the targetted adverts thing more together? For example, perhaps adverts could have a “like” and “dislike” button. That way you could mute ones you particularly dislike (a win for you), the advertiser at least gets some feedback (a win for them) and Spotify get data which will enable them to make better guesses about what fits where and and also a tool to measure that (a win for them and probably a PhD thesis for someone in information science, too!).”

    That’s a really valid point. That being said, when offered the opportunity to express oneself on whether or not they like an ad, most people would say they dislike them, even if they are fairly targeted.
    So I’m not sure that would really work.
    The problem is the targeting. When you advertise on the internet, you’re generally pitching to a specific demographic. If you have an ad on a women’s magazine website, chances are you’re selling a product that you believe could be of interest to women, and in most cases, you have a fairly good idea of the age group, social status, etc,

    When you put an ad on the radio, it’s a similar story. The radio had data on the demographic of their listeners.
    On radio stations that focus mainly on music, a record label can advertise by plugging a song. On less music focused radio stations, the advertising is geared more towards other things. ie: you’ll hear an ad for a bank just before a programme that specialises in finance, etc…

    But Spotify is not a website, and it’s not a radio station. It’s an on demand music service, where its users choose what they want to listen to. It’s not a passive experience. You choose an album you want to listen to as opposed to just listening to what the radio station has to offer. So how do you target ads? Can you determine a person’s demographic just by analysing their taste in music?

    Spotify only knows my postcode, the country I live in, my age and whether I’m male of female. So they can try and sell me men’s underpants if they want!

    Another problem faced by Spotify is the production and implementation of ads. If you advertise on the internet, you can go through Google AdSense, or similar advertising services, based on keywords, etc… Everything is text based. And the normal website ads generally include text, images and in some cases sound. You can choose what kind of website your ads appear on.

    I don’t know about you, but I don’t look at the Spotify application very often so the only way of grabbing my attention is by playing an audio ad. So it seems to me, Spotify is missing out on a large amount of potential ads because their medium doesn’t fit very well with the current advertising models. So you’ll hear crappy homemade ads or see a large takeover in the application. Based on the quality of some of the ads, it’s pretty clear the advertisers didn’t spend much money on production, so it seems pretty obvious that they didn’t spend much money on their advertising campaign either.

    If they want to start making some money without annoying people, they should do something like powerplay or Jango Airplay. You choose a bunch of artists that are similar to your bands, and you pay to have your music played to people who listen to those artists. It solves the recommendation issues in Spotify (to a certain extent) and will make Spotify some money in the process.

    Another idea I had for Spotify is for them to build a custom platform for ads that changes the background music of the ad based on your taste in music. A lot of the ads don’t even have background music and sound terrible.
    If you do a mash up of both ideas, what you get is a less disruptive ad because you move seamlessly from the music you like to music you probably like with an ad sitting on top of it and an opportunity for unknown bands to provide their music for an ad in exchange for a some exposure. In this case, the band would just provide the music for free of course.

    Spotify wouldn’t even have to worry about how to choose the music because the bands providing their music would select artists they think they sound like and the targeting could be done based on that.

    It doesn’t solve the problem of whether or not the contents of the ad is well targeted or not, but it would certainly improve the user experience and increase their advertising revenues.

  4. If spotify is aiming at paying similar rate to artists as radio does per listener then presuming advertisers will pay similar rate per listener as radio ( and well targeted adds should be a premium product) there is no reason why this model shouldn’t be made to work. If its currently loss making thats spotify’s prob not mine, and if artists are getting similar rate to radio then i don’t see it as their prob either. Far as premium goes – i can see how an add free product could develop into an alternative to cd buying so can see artist case for higher rates – on the other hand the maths rather depends on how many times you’d expect fans to listen to an album if they bought it. ( And how much you’d have expected as an artist from each sale) I can also see a strong case that a play is a play is and my choice to listen to adds or or pay to remove them is and irrelevant to how much the band should be paid. Sounds like the royalty rates need very close examining, which spotify seem to be doing their best to avoid. At the mo this isn’t really a prob spotify is small and if its replacing cd purchases for only a handfull of folks, but if we moved to a future where streaming were the norm rather than ownership then the precise details become a vitally important issue.
    We need transparency from spotify pretty soon.
    Oh in calculating how much artists should get per spotify plays i’d best point out half my plays i have on cd just can’t be arsed having a rumage in my boxes of cd s.

  5. “Please explain the royalty structure if it’s not what I said it was.”

    No. I can’t explain it in detail. All I can say is that premium plays do count.

    Count in what way? That doesn’t help. If you don’t know, say so. If you do know, let us know how. My initial statement was based on a quote from Daniel at SXSW. If it was wrong, as I’ve said, that changes everything. So far you’re not giving me anything to go on.

    “As for the Ads – I’ve NEVER heard or seen an ad that was useful.”

    Is it the same if you watch TV? I have only used Spotify free for short periods of time and in those times there were many very useful ads.

    I don’t own or watch TV. When I do watch TV programs via the web, it’s on the BBC iPlayer, so no ads.

    “If Spotify plans to fund it self on a 60/40 split, how close is it to that now? Do you know? Will you tell us? Do you know someone who will?”

    I don’t know. My guess is that they have a lot of work to do in the UK and Spain, but doing great in the other launch countries.

    Based on what?

    “I’m happy to be wrong – I want these things to work. I do. I don’t hate Spotify. I *may* pull my music off the service, just to see what happens. I don’t need it, it doesn’t need me…”

    How many listeners do you have in Spotify and why would you even consider hurting them? Do you really want to anger your fans by actually pulling your music off Spotify?

    My music not being on Spotify isn’t going to ‘hurt’ anyone. It may mean that some people come here to download it for free instead (Because they can). How many listeners do I have there? No idea. There’s no way of knowing. Spotify don’t publish any such data. Unlike, say,…

    As of today I haven’t seen any stats indicating that Spotify is taking anything away from artists. There is no cannibalization and yet you and many other artists act as if it is a fact. I find it both strange and frightening to see small artists think exactly like the very big ones in this matter. If you aren’t losing money, but rather making money, what is so bad about being there? If you have stats to show that you actually lose money because of Spotify it would be great to see them!

    Now who doesn’t understand. You’ve been linking to my post about the economics of Spotify for artists for ages – I don’t think Spotify is remotely ‘bad’ for artists. I think, as I say in this post, that Spotify has failed to differentiate between Spotify lite and Spotify premium in a way that makes sense to me. They’ve broken a link that I see as vital between payment and expression of value in and gratitude for music. At the moment they don’t have anywhere near enough premium users for it to be a big issue. Largely because £120 a year is WAY more than 95% of music listeners have ever spent on music. That’s a lot of money going into the music economy. The big question is how it comes out.

    There are loads of big questions over who’s getting it – every single bit of info I’ve seen about Spotify rates/payments/etc. has been speculative because they won’t/haven’t published any figures. Given that they are making money from playing adverts to people who are there listening to my music, it would make sense for them to tell me how much I’m getting, and how it compares to the rest of the artist there. That would seem to me like a fair, honest and open thing to do.

    So *if* I do pull my music off Spotify, it’ll be for one of two reasons – a) because I’ve decided that I have an issue with ad-funded music services all round (a much bigger consideration that I need to think harder about) or b) just to see if I notice any impact… see if anyone bothers to contact me about it, see if I see a drop in plays/sales/etc. I’m willing to experiment with my own music career so that musicians can find out this stuff. At the moment, you and I speculating about what’s ‘good’ for musicians/music without any hard data on either side (unless you’re privvy to hard data but are unwilling to post it, in which case, I’d appreciate it if you’d stop commenting until you are willing to share it…) isn’t really getting to the bottom of it…

    1. …I’m guessing you’re playing blog catch up, cos the next post after this one is a response to that very post 🙂

  6. Spotify cancelled my Premium account for me!

    Will blog but also has something to add re: their advertising models and privacy and an email conversation I’ve had with their legal council.

    Stay tuned people.

Comments are closed.

© 2008 Steve Lawson and developed by Pretentia. | login