To stream or not to stream?

A lot gets written about streaming music. It isn’t all true. Perhaps even more frustratingly when it is true, it isn’t always the whole truth. Here are ten misconceptions about music streaming, which I’ve attempted to explain. If you have questions, please use the comments section, and I’ll do my best to answer them or, at least, to explain why I can’t.

1) Streaming services are all basically the same

Streaming services are not all the same. They all play you music over the Internet, but that doesn’t make them the same. They use different technologies, have different content, different ways to find music and they don’t even all sound the same. We run four different streaming services at price points from $19.95/year to $300/year per subscriber. There are lots of others.

2) Royalty rates are scandalously low

There’s a big range of royalty rates paid by streaming services. We have deals with services including Rhapsody, Napster and Spotify, but we also have our own streaming services (Naxos Music Library, ClassicsOnline and so we get to see a lot of the figures. Services pay us between three cents a track and a quarter of a cent per track1.

Whether or not this is low depends on what you’re comparing it to. If your reference point is a paid download2, then yes, it’ll seem low, but the two simply aren’t comparable. A stream is ephemeral, a download is permanent. In this regard, a stream has much more in common with radio than it does with either a download or a CD.

When one of our tracks is played on internet radio in the US, we get about a quarter of a cent per listener3, comparable with the very lowest streaming rates. We get a little over half that from satellite radio, and nothing at all from traditional radio.

The royalty rate correlates closely to the degree of interactivity. At one end of the spectrum, you’ve got the download, which is yours to play whenever you want. At the other end is traditional radio, where fixed playlists and a small number of stations give you very little choice indeed.

This correlation even applies among streaming services, where Naxos Music Library (an academic research tool designed for people who really need to hear a specific track in detail) pays a higher rate than consumer-focussed sites, which are more likely to be used for background music.


3) Streaming services are not for serious listening

This accusation has been levelled at just about every innovation since letting common people into the concert hall. The gramophone reduced serious symphonic works to 7-minute soundbites. LPs lacked the clarity and immediacy of shellac. CDs sanitised the sound and took the tactile ceremony out of listening. The iPod reduced serious symphonic works to 7-minute soundbites. Streaming made us careless consumers who don’t really pay attention to what we’re listening to.

The idea4 is that if you don’t have to pay for music before you listen to it, you don’t have much of an investment in it. You’ll pick the singles, skim over the album tracks, and get stuck in an ever decreasing circle of superficial instant gratification.

It’s a plausible argument, but it’s the answer to the wrong question. We’re not looking at a world in which streaming subscriptions become the only way of consuming music—just a world where they’re one of many options. Streaming has been around for about a decade, and it shows no sign of taking over. People still press vinyl, CDs are very popular and downloads have a growing share of the market because they all have different advantages, and each is the best solution to a different problem. The same goes for streaming.

It’s true that streaming services mean you don’t have to make a financial investment in a piece of music before you hear it, but there are foreseeable positives to this too: with nothing at risk but your time, you can take a chance on new repertoire.

Public libraries didn’t turn out to be bad for literacy in the long run. In the same way, we think the wider availability of music will be good for music appreciation.

4) Streaming is bad for artists and/or independent labels

We’re adults. We look carefully at each contract before we sign it5. If we thought streaming subscription services were bad for labels, we wouldn’t have started our own and we certainly wouldn’t have signed up with others.

There are some widely-reported examples of artists receiving very low payments from streaming services6. It’s worth remembering these payments are made under the artist’s contract with the label, to which the artist agreed as a consenting adult, and that contract may involve the deduction of a proportion of those royalties. You can’t work out what the service is paying the label by looking at what the label pays the artist.

5) It’s all a scam cooked up by the major labels

It’s not a scam. It’s a deal. Consumers should do it if it makes sense to them, labels should do it if it makes sense to them, artists should do it if it makes sense to them. If a lot of people do it, it’s because a lot of people think it makes sense. If it doesn’t make sense to you, don’t do it.

6) It’s the end of the record collection

As a consumer, if you like records, there’s nothing stopping you from buying them. Almost all releases are still available on CD and as downloads. Vinyl is growing in popularity.

For artists and labels, there’s not a lot of evidence to suggest streaming leads to an overall reduction in the sales of other formats. There’s plenty of speculation about how consumer behaviour will change, but the only way to find out for sure is to try it and see what happens. So far, from where we’re standing, it looks good.

Streaming services certainly do change our relationship with the recordings we hear. Unless you’re very young, it’s likely you already own a whole bunch of music, so a service like Spotify or Naxos Music Library becomes an addition to the collection.

For some people, it’s a way to check out recordings before buying them as downloads or CDs. For others, it’s a way to hear the music they don’t like enough to buy. Some might stop listening to their existing collections altogether, choosing instead to create playlists of favourite albums online. No matter how we use streaming services, we’re still actively involved in listening, and in curating a selection of music. It doesn’t look the same, and it doesn’t take up anywhere near as much space, but all the musically significant parts of building a record collection are there.

7) You’ll be locked in

If you spent your life putting together your record collection—if you’ve carted crates and crates of LPs and CDs from your parents house to college to your first flat to your first house to the place you raised your kids, adding all the time to a library of physical objects that together form a soundtrack to every special memory you possess—if this sounds like your collection, you might well be uneasy at the idea that your music might no longer belong to you, and that you’ll lose it if you stop paying for it at a price that could increase significantly at some point in the future.

This is, though, only half the story. You’ve still got your records, and a subscription isn’t going to take them away. Unlike a physical collection, any new playlists you create are stored on remote secure servers, can be easily copied, and will survive fire, flood or burglary in your home. Right now, a Spotify playlist isn’t compatible with Rhapsody, but as the market matures, it will become easier and easier to transport the playlist element of your collection from one service to another.

8) Successful pop artists are holding back releases from streaming

“Because Coldplay does it” is not a good reason to do something. If it was, I’d have a child called “Apple”.

In the film industry, it’s normal to release a film first in cinemas, then for rental, then to buy on dvd, and finally for broadcast on television, giving each platform a chance to play to its strengths. This works well for blockbusters, but smaller movies adapt the pattern to suit their own niches.

We may one day see a similar standard release pattern for music, although it’s unlikely we’ll see one plan work for everything.

In the meantime, you can expect to see more albums held back as our whole industry tries to work out what the ideal pattern looks like for each type of music.

9) Streaming services aren’t sustainable. They’re just trying to build interest so they can sell to Google (or somebody else).

Articles like this one contain a lot of speculation. I can’t speak for anybody else’s business, but I do know that Naxos Music Library is already profitable and continues to grow.

10) The audio quality is bad.

If this is your only objection, then you want to be wrong about this, because this one issue stands between you and great convenience. Research has repeatedly shown that our expectations have a big impact on our perception of the music we hear, so do yourself a favour and enlist a friend to help you with a blind trial. I suggest comparing Spotify on the premium subscription (which streams at up to 320 kbps) with a CD (or WAV/FLAC files) of the same music. Have your friend test you: if you can’t tell one from the other without peeking, sit back and enjoy the music.



1 Streaming services typically pay out a fixed percentage of their revenue, shared out amongst the labels according to the proportion of total streams apportioned to that label’s content, so if your music was responsible for ten percent of total listening, you get ten percent of the total money. When a service has just launched, most users will be on a free trail, and the pre-stream revenue can be very low. Sometimes there are per-stream minimums built into the deal, but these don’t always determine the sum you actually get, and it isn’t the same every month, so it would be misleading to simply say “we get x per stream.”

2 When you buy a track on iTunes in the US for 99 cents, we get 70 cents. We pay 9.1 cents in music publishing (if the work is still in copyright) and that leaves us with about 60 cents, which is somewhere between 20 and 300 times what we might get from a stream. It’s tempting, then, to say that unless users listen to their downloads an average of 20-300 times, they’d be better off with streaming and we’d be better off with downloads. This is an oversimplification: it assumes that users only stream or download, that nobody buys anything after streaming it, and that streamers and downloaders listen to the same amount of music, regardless of cost.

3 More on US rates for digital radio here:

4 Gabriel Kahane puts the case eloquently here:

5 The deal terms described in this article are fairly accurate. What isn’t accurate is the analysis, which misunderstands the nature of a monopoly and wrongly presents the terms as onerous, unfair, or impossible to deal with.

6 This article describes one example:

Leave a Reply