Spotiwhy? : Are Subscription Music Services a Sustainable Business Model? | Hacker News Reader

Spotiwhy? : Are Subscription Music Services a Sustainable Business Model?

musicthinktank.com

46 pointsOoTheNigerian31 comments

So the author says they are, up to a point, and assuming they get significant penetration. But it's based on an average listening rate of 2.5 hours a day, which is apparently what an average american teenager listens to. I'd have thought the average american teenager probably has the highest music listening rate of any demographic. I know my own usage would be closer to an hour on the days I queue up music, but that would be only every other day at best.

If that is closer to the average listening habits, then the royalty per stream calculations end up over 5x more profitable and above the magical 1 penny/play mark.

I would also argue that a big part of listening happens on the move: listening to music on mp3 player or phone.

And that's a separate products that can probably be sold for separate money.

Yes I would expect a long tail of listeners, imagine I am at less than an hour mostly. To some extent this will just affect penetration though, as the light listeners will not pay $10. In principle you could sell them cheaper capped play subscription though.

On the other hand, he counts 2.5 hours per household. Although Spotify does not allow for simultaneous plays using one account, more than one person in a household may use one account.

Also, the article assumes the premium subscription rate, but probably very many users have the 'Unlimited' subscription. So, although the average is probably lower than 2.5 hours, the average subscription price is also.

Not necessarily. I'm a programmer (not a teenager anymore), and I listen to music almost all the time while working, which is definitely more than 2.5h per day. Of course I might be an outlier, but as streaming gets more popular and easier to use, more people would listen to it in the background, like radio.

Thats what I thought of, but then I looked up average radio listening hours, and they seemed to be about 0.2 hours per person (UK figures). So I think you may still be an outlier.

I do have to admit, if it weren't for work or walks, I really couldn't care about listening to music.

Yes, but, radio listening hours? Probably the under-30 demographic only listens to the radio under duress. My knowledge of this is anecdotal only, but other than w/r/t exceptional cases (NPR, highly curated college stations), there's perceived generally to be near zero incentive to listen to the radio by everyone in my peer group or behind my peer group in generational terms (I'm thirty). On the other hand, those same people seem more or less wedded to other forms of consumption (digital music players, pandora, etc). Listening when exercising, listening in the car, listening when appropriate at work, listening while cooking dinner, listening during coitus, etc. etc.

Very nice breakdown and analysis of the business model.

I too agree that assuming 2.5 hours per day on average is a stretch. I generally listen to more than that, but i'm fortunate enough to have a job where I listen to music while working. Most people don't get to do that very much.

Subscription music services are my preferred way of consuming music. I had a Zune subscription for over two years (eventually cancelled when I lost my Zune and just started using my iPhone... Zune's big weakness was its hardware exclusivity), and I signed up for Spotify premium as soon as I could. I like sub models because I am paying for music, I can get whatever music I want (almost... bands like Arcade FIre don't license their music to these services I've found) whenever I want, and the actual data is well-manicured in terms of metadata.

Rdio has Arcade Fire. Strange that Spotify doesn't....

But yeah, I agree, subscription music services are the future. The holdouts (Beatles, Metallica, AC/DC, etc.) will eventually license their stuff, otherwise their fan base will erode. I already think that their tardiness to the digital download party has hurt them already.

Subscription music services are as sustainable as the content owners want them to be. Just look at Netflix. The content owners take advantage of the tech and foothold providers have, then eventually squeeze the providers until their own distribution system comes online.

If you don't consider it that way yet, it's coming. The bright side is that eventually "content owners" stop being the major labels and become the actual musicians. This industry is ripe for a wide open distribution market.

"The bright side is that eventually "content owners" stop being the major labels and become the actual musicians."

I keep hearing this, but I still don't buy it. Record labels do actually provide a service to artists- generating publicity, scheduling live performances, etc. Don't get me wrong, a lot of them are evil, but an artist is still at a distinct disadvantage if they are labelless. There's a reason why only bands like Radiohead made a splash when they went without.

I don't think the argument lies with whether or not record labels provide a valuable service to their artists, they do, but more if the value of their services will remain above their cost. When they were the only distribution game in town, it was a no-brainer as you either distributed through them or you didn't distribute on any meaningful level at all. Now that there are more and more tools and services available to artists that replace some, and possibly soon all of the labels' functionality, the question becomes at what point does their cost outweigh their value?

With distribution through itunes & amazon I wonder just how long the record labels' business model will be sustainable.

90%+ of the music I purchase is self published. (and incidentally, from searches I've asked friends to make, it appears most of what I listen to is missing from spotify.)

Cost of goods sold for an artist are fixed whether their song is streamed 100 times or 100 millions times. After that cost is recouped, it shouldn't matter much what the rate paid to the artist per stream is. Why are we still clinging to this expectation that artists must be multimillionaires to be satisfied with their careers?

I was a Rhapsody user from its inception, maintained an account even when I bought a Zune and got a Zune Pass and now switched the Rdio + Spotify and bought an iPod touch. For me, I consume a lot of music and I want the easiest way to sync my collection to the devices I use -- office, home, portable -- and so, I'll happily pay for that. The alternative before was simply to download it and pay nothing. Physical music is nice but it takes up space, is prone to being lost (I once lost 200+ CDs on a transcontinental flight in my late teens...it sucked. A lot) and I'd prefer the cost of 2 albums a month to hear whatever I want over the cost of buying the media. I'd consume a lot less the other way.

The fact these services have finally hit the mainstream is no accident.

Ok, but spotify is available in the following countries as well: Sweden, Norway, Finland, the UK, the US, France, Spain and the Netherlands.

Why not use those as well to calculate the possible yearly revenue?

You can do the numbers... The article is about how to model this. Expandi g to the whole world does not change the argument much as you get a bigger pool of artists too. He was comparing off US revenues on eg CD sales. There are still huge national differences in music preferences.

I've seen a few discussions of Spotify revenue models, but they all miss the most obvious (to me, at least) comparison: That with radio, which seems to be the closest match. What's the revenue per play per listener there? I'd expect Spotify to compare rather favourably with those.

Radio doesn't kill the necessity of owning music the same way as a subscription service. You can't listen to your favorite 10 love songs over and over after a break up on radio. Or the band that your peer group has elevated to exaltation. Radio can plant a song in your brain and you then want to hear it more often thus making you more likely to buy your favorites.

A related infographic from last year by David McCandless, "How much do music artists earn online?": http://www.informationisbeautiful.net/2010/how-much-do-music...

The author makes three errors, one of which he stipulates (the he doesn't know the revenue split). The other two are:

1. Streaming is basically a perpetual source of revenue. If 1 download is equivalent to 200 streams then ask yourself how many times on average does someone listen to a purchased song?

2. The comparison you should be making isn't to digital music purchasing but to radio royalties. The formulae for this are complicated (flat fees to BMI, ASCAP, etc and then a complex formula after that for distribution to writers but currently not performers, except for Internet "radio").

Of course streaming (renting) earns you less than a purchase.

The recording industry likes to paint this as the artists are getting screwed. And they are getting screwed but not by streaming services. They're getting screwed by the recording industry [1].

Fact is, very few artists can survive on CD sales and digital downloads alone. Most artists survive by performing and even that's being eroded by the recording industry who uses fall guys like Ticketmaster to charge "fees" that fall outside of the revenue split with the artist on ticket sales. The public hates Ticketmaster. Artists hate Ticketmaster. The industry loves Ticketmaster.

[1]: http://www.techdirt.com/articles/20100712/23482610186.shtml

> If 1 download is equivalent to 200 streams then ask yourself how many times on average does someone listen to a purchased song?

For most downloaded songs, a couple of times.

Most people just amass mp3 collections they rarely listen to in their entirety.

A couple times?? Surely some people do that, but I doubt it's the majority. This is way too broad of a generalization.

I'd say, anecdotally from my own experience, that it's more of a power law distribution. Some songs/albums I listen to quite frequently, while others I listen to once and then forget about.

Suffice to say using this sort of subscription model, the hits would still be big money spinners, but for the vast majority - the long tail so to speak - wouldn't make enough to live on.

How does the music business differ from the book business? Authors do not plan on supporting themselves by touring. Why is this more or less expected for musicians?

Are authors worse off for not having a viable touring option?

I'm mainly curious about assorted business models that work well for musicians for whom live performances are not practical. If one person with a pc and some electronics makes an album, live gigs may be nonstarter.

In some regards it's even worse for authors. A lot of authors, especially of technical non-fiction books, have day jobs.

Regardless, the reason why this is so is simply because it has become the norm. There is nothing about it that is particularly good except the degree to which it concentrates wealth in the hands of publishers (which they certainly enjoy). Over the long term I expect the norm will change to self-publishing or very low-overhead mostly self-service publishing.

> The recording industry likes to paint this as the artists are getting screwed. And they are getting screwed but not by streaming services. They're getting screwed by the recording industry [1].

This isn't an or proposition. Artists are being screwed by both. Yes, the industry takes the lion's share of revenue from artists. That doesn't mean that on-demand streaming services aren't a joke in terms of compensating artists.

And you're wrong about your #2. The comparison should not be made to radio royalties because radio play has been shown to increase sales, since if you want to hear a specific song you heard on the radio, you still have to buy it or sit around waiting for it to come on. People compare Spotify revenue to purchasing because, if your music is on Spotify, there is no reason for anyone to buy it anymore. If it replaces purchasing for most people, the revenue needs to stack up by comparison.

That doesn't invalidate my point because curation, which you can argue radio is, and promotion (radio stations are paid to pay particular songs) is still possible with streaming. Just look at Pandora, last.fm, etc.

Also there are different definitions of streaming. The purest definition is where you have to have an Internet connection to listen. There will be various levels of caching involved here. For example, the service might cache the entire song--possibly several songs--to avoid stuttering and service interruptions.

Another model might better be called "cloud management" of music where you have a cloud-like iTunes interface (or even a non-sucky interface) that would sync that playlist to your device for offline listening, something I think iCloud will either be or quickly become.

Bandwidth and service issues are going to make local storage of music a superior solution for some time to come (IMHO).

I wasn't arguing for or against the article or your overall argument. I was simply pointing out some issues with two of the things you said.

Reply on news.ycombinator.com