18 Oct 2015

How Much is a Stream Actually Worth?

Last week, the UK's Music Managers Forum published a report "Dissecting The Digital Dollar", which "set out to explain [...] how digital income is shared between each stakeholder in the wider music industry". It is a good report, in which they outline seven key issues and raised 15 questions for the industry to think about.

But they failed to include the most important issue: The value of a single stream. Without that point, all further discussion is rather meaningless. Whatever percentage is applied to nothing, it'll still be nothing.

The music industry will need to consider the per-stream value that any percentages are going to be applied to. The MMF report totally fails to address that point, and puts the "Division of streaming revenue" at the top of their list of issues.

So let's look at a number of different price points that are currently floating around the industry:

1) Same as a download?

On the one extreme, we still have people who compare per-stream rates to download prices. But streams are to downloads as rental is to purchase. The monthly rental value of a house is only a fraction of the purchase price of that same house. The daily rental value of a car is only a fraction of the purchase price of that same car. But property owners make ends meet, because they can rent the property month after month. Car rental companies make ends meet, because they can rent the car out day after day.

A stream is a short-term rental of a track, whereby the rental term equates to the duration of that track. When that user wants to listen to that track for a second time, it counts as another stream which results in another payment to the rightsholders of that track.

The value of a single stream ("rental") is therefore only a fraction of the price of a download ("purchase"). Basing expectation on 79p/99c per stream is unrealistic.

2) As much as advertisers are willing to pay?

On the other extreme are the providers of ad-funded services. I have no issue with the idea of ad-funding. Ad-funded models can work. Ad-funding works fine with commerical radio, because advertisers on commerical radio are willing to pay enough to fund the radio stations' costs for the music they play.

But it appears that in the streaming world, advertisers are not willing to pay enough to satisfy the rightsholders, artists and writers. Should non-music companies such as Coca-Cola, Ford and McDonalds determine the value of music? No! It is up to the music industry to decide that value, and if an ad-funded business model's advertisers are not willing to pay enough to cover that value, then that business model is not going to work. Such a model is simply broken.

A lot of ad-funded streams earn absolutely zero. This happens when an ad-funded service has no "advertising inventory" available to sell against a particular stream for a particular user at a particular time. But that should not mean that the stream has no value. Somebody is listening to the track, and (hopefully) getting some enjoyment out of that. Rightsholders should get an amount of money for that, irrespective of "available advertising inventory".

The music industry can enforce that by setting an absolute minimum rate for an individual stream. To make ad-funded models viable, the services may need to sell more advertising and/or rise their advertising rates. If they then can't afford the minimum rate, then they don't have a viable business model.

They are supposed to be "ad-funded music services", but some behave more like "music-funded advertising networks".

3) A penny per track?

Somebody recently pointed me in the direction of streaming service with an alternative model. That service follows a "metered" approach: The user buys an amount of credit, and then each stream uses 1p of that credit. When somebody buys £5 credit, he can listen to 500 streams, and then needs to buy more credit to keep on listening to more music.

This is a much more reasonable approach, as it takes into account how much a consumer is willing to pay for the product.

Whether it is enough for the rightsholder is a different question. The 1p includes VAT, and (under the current 55/15/30 split) a label would thus get 55% of 0.83p per stream, which is 0.46p. That is probably not quite enough for many people in the industry, but it is already significantly better than the current rates from ad-funded models.

4) As much as the subscribers are willing to pay?

In the early days of mobile phones, users would get charged on a per call basis. Then, about a decade ago, the mobile phone companies worked out that consumers are willing to pay more for subscriptions, bundles and unlimited plans. They found that users on a subscription will make far more and longer calls, and are willing to pay a higher price for that, often to the level of a higher per-unit price.

The same principle has already been proven in the world of streaming music. Users on a subscription end up paying more than the 1p per track from the previous metered service. There is sufficient data floating around to show that an average premium subscriber to a streaming service listens to around 500 to 600 streams per month. Do the maths and you will see that such a premium subscriber pays more than 1p per stream. (Don't forget the VAT in the calculation).

Is it enough? That is still open for discussion, but at least this price point is purely based on what consumers are willing to pay for music.


There is a big variation in the value people put on a single stream. For me, the per-stream values from paid-for subscription services are acceptable. But the per-stream values from purely ad-funded services (and from ad-funded tiers in freemium models) are much much lower, often to a ridiculously low level.

Sorting out that discrepancy should be the music industry's top "Key Issue".