The average musician gets $23 for every $1,000 of music sold


I used to write a lot about the music industry but it because less interesting once Spotify became dominant and everyone accepted that streaming is the future. This chart got me excited again though. Pop stars regularly complain about how little they get when their music is streamed, but they are mistaken in blaming the streaming services. It’s the labels that have the biggest take.

The $1,000 we are talking about here comes from music sales – e.g. CDs. As you can see the label takes $630 from that $1,000 and the band gets $230 which is then shared with their advisors leaving band members in a typical four person band with $23 each.

I saw the chart on Techdirt, a site dedicated to exposing old media rip-offs and BS. It’s great to see them still going. After the chart they go on to explain how even $23 is recoupable against any advance the band might have had, so the true situation is even worse.


Good middlemen in most modern industries have a 10-20% stake. AriBnB, eBay,, TripAdvisor and many others are in this range. Companies that take a 63% cut are open to disruption and it’s hard to see how the labels have held onto their position, much less how they will sustain it.

  • The numbers in book publishing can be similar.

  • All the old media industries are the same! That’s part of why we backed

  • Rob Shreeve

    I’m a huge fan of your blog Nic. I am not employed by a music label or a publisher, (though I was a publisher). But to equate the activities of a music label or a book publisher (Jon) to airbnb and eBay is utter nonsense. The latter are marketing platforms for other peoples products – they take a cut from selling the fruits entirely of other people’s labours. They don’t take financial risk in signing up and paying, often large, advances to creators. They don’t add value through skilled curation, shaping, design, production and promotion. (I know airbnb do market but on an aggregate level). Nor do they get stuck with the product if it doesn’t sell enough to cover costs (which over 50% do not). Having said that the music labels do take too big a share (of digital sales especially) and book publishers are only gradually revising their terms for eBooks. But to say that they’re like Airbnb and should therefore only take 10% is a bit of a nonsense. It rather devalues an otherwise fascinating set of stats. (fyi Jon 50% of the price of any book goes to the retailer.Another 17-20% is cost of production. 8-10% is warehousing and distribution. then 8-10% to authors. Do the maths…

  • Hey Rob – thanks for reading and don’t worry about calling me out on stuff. It’s how I learn.

    Here’s where I see the parallel with AriBnB – you could have an online marketplace that bands use to distribute their music. It could have the following features:
    – helps consumers find new music
    – is the best place for them to access the bands they already know
    – curated music – top ten lists, curated radio etc
    – personalised music feeds
    – social feeds – what my friends have been listening to and/or what other people like me have been listening to

    All that could be done for 10-20%.

    Two major things labels do that are missing from that list is advances and above the line marketing campaigns. They definitely cost money, but I don’t think they are worth 40-50% of sales.

    You make a good point though, in that somebody needs to finance initial production and that needs to be part of the model.

    Our portfolio company Unbound does something similar for authors, charging cost for editing, printing etc. and then going 50-50 on the profits after that.

    Make sense?

  • Rob Shreeve

    Just seen your reply Nic. What you’re describing sounds like Spotify with a sales channel. But that only makes sense if as a buyer I can find everything they want there and that needs a comprehensive catalogue. There have been several abortive attempts to provide a platform for new, un-signed bands but they’ve all failed. Publishers’ and labels own sites failed in the early days for the same reason.So we end up with comprehensive ranges from aggregating retailers like Amazon who charge labels and publishers 50% +. They dont curate (apart from their current direct publishing attempts to dis-intermediate publishers) so we’re left with the need for someone to find new authors and acts, develop them, handhold, be their best friend and confidente in their darkest moments of self doubt, take a financial punt on the first, second or third releases before they make a penny and spoon feed the retailers with returnable stock at thumping discounts. It’s not an easy business.

    As for Unbound, I like its business but they’ve found a smart way (Crowdfunding and pre-selling to committed buyers) that cuts out the risk element and guarantees a viable print run and a profitable sale. It’s the Holy Grail for publishers and they try to achieve that by publishing bankable authors with guaranteed fan bases who WILL buy. But to get, or hold onto, those authors and artists they have to pay very large advances and pray that they’ll earn that cash back. Unbound, is great for unknown authors with a saleable idea or superstars who want to sell direct to their tribe, But that’s unlikely to work for the 100,000’s of new books published every year. For the rest (90%+ I’d argue) you need curation, editorial expertise, market knowledge, passionate advocacy and consistency PLUS marketing muscle and the cash to take risks.Without them thousands of books wouldn’t get published and hundreds of artists would remain unknown to anyone but a niche audience.

    To follow your airbnb analogy: I doubt very much if airbnb are going to start buying houses. Uber may have plans to create a fleet of driverless cars but I’d lay money that they’ll make sure others take the financial risk of buying them.

    So interesting debate and far longer than I intended for which apologies….


  • Interesting. To pick out one point – Unbound is “good for unknown authors with a saleable idea” – true, but the hope is that over time they can will be able to scale the model down to authors that whose idea is only saleable in a small way. Arguably the rest should just self publish…