After a few years of experience with streaming music services, and quite a few impressive analyses of the economics involved, a consensus seems to be forming around two points: (1) the current models for streaming services are unsustainable; and (2) the next generation of streaming services will look very different from the current generation, but no one can say what they will actually look like.

The first proposition appears to be true almost by definition. Various analysts have persuasively shown from every angle that ad-based streaming and subscription-based streaming cannot make money in the long run (or even in the short run).  The economics simply do not work, no matter how far you scale the customer base.  In fact, it appears that the more the services are scaled, the more money they lose.  Simply put, it appears that these are simply not sustainable business models.  

The second proposition seems to be half-true, half-false.  Of course it’s true that the next generation of streaming services will look nothing like the current generation, because that transition will occur in order to replace unworkable models with potentially workable ones.  It would make no sense to replace one generation with another that is exactly the same.  It is not true, however, we have no idea what the replacement model will look like.  In fact, there appears to be a very strong contender for that next model, with an equally strong rationale.  

So why are the analysts and experts not recognizing what that next model looks like?  The problem, it seems, is that everyone writing about subscription models tends to ask the wrong question. They ask “how do we price the ‘unlimited everything’ subscription model to make it work?”  

For example, Ted Cohen said to HypeBot:

But then we have recent analysis from Rockonomic suggesting that the right price is around $3.99.^

If you read assorted commentary about the tribulations of Spotify, you will find multiple variations on this theme.  The price should be higher, to maximize revenue.  No, it should be lower, to maximize revenue or to keep from shutting out fans, or whatever.  Everyone has a personal opinion about what the magic subscription price should be (or should not be) to solve the problem.

But in truth, this is probably not an argument worth having.  In the end, it doesn’t really matter what I think, or what you think, music is worth.  If I think it’s worth $5, and someone else thinks it’s worth “more than $10,” who is right?  Who is wrong?  As long as we’re each spending our own money, we are both right—and, more importantly, we are both completely wrong.

It boils down to the fundamentals of economics:  Basically, something is worth exactly what someone will pay for it, no more and no less.  Any approach to pricing streaming services that ignores this law by definition prices the product incorrectly, excludes consumers, and leaves money on the table.  More astonishingly, pricing a streaming subscription in this way eliminates one of the prime advantages of digital delivery of creative content: the ability to customize that delivery to each and every consumer, without regard to what other consumers are doing.

 

Let’s shorthand a quick example.**  Say we have a group of 15 consumers who want to listen to streaming music.  As is the case in the real world, each person has a maximum monthly price, based on her own experience, valuation of music, financial situation, and so on, that she is willing to pay to listen to the music she chooses.  (She may not know exactly what that price is, but it exists, and is identified when she is presented with a yes-or-no option to subscribe to a streaming service—she either pays the set price or she pays nothing and gets no music.)

 

Further, let’s say our 15 consumers break down as follows:

5 people will pay a maximum of $0

4 people will pay a maximum of $4

3 people will pay a maximum of $7

2 people will pay a maximum of $10

1 person will pay a maximum of $12

The consumers, in total, are willing to spend $69 for music.  Now, let’s ask the question that this subscription pricing debate always raises:  

No matter what answer is chosen, it’s wrong.  Why?  Because any price named leaves money on the table (out of the $69 these consumers would pay in total for the service) and therefore does not maximize revenue. 

For example, if you use Spotify’s $10 price, you get 3 subscribers paying $30 … but you miss out on $16 from the $4 crowd, $21 from the $7 folks, and $2 from the $12 person, who pays less than she is willing to. Spotify leaves $39 on the table, out of a total of $69.  That’s more than half!

Fine, you say, we’ll go to $7.  That’s not much better, because we’re still getting only $42 out of the $69.  Sure, we got the $7 folks on board, but we forfeited $6 from the $10 crowd and an additional $3 from the $12 person.  And we still missed out on $16 from the $4 group. 

Dropping to $4 doesn’t solve the problem, either.  Now we’ve got all 10 people who are willing to pay, but they are paying only $40 total, leaving $29 on the table. 

This approach doesn’t work because any single price will, by definition, be above some buyers’ maximums and/or below others’ maximums.  Some people pay nothing because the price is too high for them, while others pay a price that’s less than they would be willing to pay.  That formula leaves money on the table no matter how you slice it. 

The simplest way to understand the shortcomings of the current subscription model is to consider our example in the context of how Spotify prices its product. 

Spotify fails to maximize revenue just like picking any single price for unlimited everything subscriptions necessarily leaves money on the table in our example.

  

Spotify doubled down on this counterproductive model by introducing a $5 tier for college students and a $1 tier for a three-month introductory period.  It is undeniable that Spotify’s per-user value is trending downward, and that Spotify is leaving an untold amount of money on the table, as every person willing to pay $10 is able to get essentially the same product for which they would pay that much for far less.  Some of those losses are recouped by getting people in at lower prices who would not pay $10, but the multi-tiered structure is undoubtedly costing Spotify significant revenues.  

It doesn’t matter whether you listen to Ted or Rockonomic and increase or decrease the price in the question to try to make Spotify’s model work: as long as you stick to Spotify’s model, you lose.

Some would argue that this shortcoming is true of pricing in general.  No matter what price General Motors sets for a car, they say, it will have the same effect—some buyers will be priced out, others will pay less than their maximum price, and GM will get less revenue than is potentially available from its entire target audience.

 

But streaming music isn’t like selling cars (although Spotify and others incorrectly approach it that way).  It makes no sense to sell the various parts of a car without selling the whole car (leaving aside the question of replacement parts and very niche markets).  If someone needs a car, they need the steering wheel, all four wheels, chassis, engine, and so on.  It makes sense to offer various bundles of options (sound system, window tint, etc.), but there are some basics that simply have to be included in order for the car to serve its purpose. 

The same is obviously not true of streaming music.  There is no such thing as a “necessary bundle” of content. Each user can use exactly the content she wants, and needs no access to anything else.  “Unlimited everything” subscriptions, such as Spotify, provide essentially infinitely more content than any particular user could ever experience.  

This facet of pricing creates a counterintuitive result: being required to pay for everything, rather than only what is used, creates an impression that the buyer is overpaying, no matter where the price is set.  

The logic is inescapable: “If everything costs $10, then what I’m actually using, which is far less than everything, should cost far less than $10.”  The selling of streaming music and the selling of cars could not be more different—and yet streaming services approach their sales as though it were the same, using unnecessary and counterproductive bundling, forcing users to pay for content they will never use, and leaving huge sums of the money on the table.

Current streaming services throw away revenue in at least three ways that are obvious at first glance (using Spotify’s original $9.99 premium as an example):

1. Some people who would pay more than $10, based on their personal value and usage of music, pay that $10, rather than the higher price they would willingly pay;

2. Some people who would not pay $10 but would in fact pay something pay $0, because there is no way to pay the greater amount they would willingly pay;

3. Some people who would pay $10 for content packaged in a way that matches their actual use and value end up paying $0, solely because they believe paying for unlimited content while using limited content means overpaying.

The current streaming services leave all this money on the table because they overlook the one simple truth of selling digital content:  Revenue is maximized when each individual can buy exactly the content she desires for exactly the price she is willing to pay (i.e., different customized prices for different customized content).  

Isn’t it amazing how that one simple truth seems to have thus far eluded every player in the market?   

You simply put content out there, and individuals come along and decide whether to buy or not.  And the way to maximize revenue is breathtakingly obvious; you let each individual pay what that individual is willing to pay for the specific bundle of content they want.  That’s how you can get the whole $69 out of our 15-person crowd, and it’s the only way to do it!  

So, in our example, each individual can consume and pay up to maximum they are willing to spend. 

Our revenue can now approach the full $69, the maximum available from this group, and we have the most sustainable model possible! 

The immediate protest, of course, is that you can’t let each person pay what that person is willing to pay — or everyone will pay nothing!  You won’t get any paying subscribers, and you will have no revenue.

None of that is true.  A pay-what-you-are-willing-to-pay model is not only possible in streaming digital content, it’s the only model that can actually work.  

It’s absolutely crucial to recognize that these people are not paying different prices for the same product; each is paying a different custom-made price for a different custom-made product!  

Each has access to the same treasure-trove of content (just as on the streaming sites now), but each is paying only for that portion she actually uses, in the amounts and mixes that she individually prefers.  

Right now, the question being asked is “how do we make the most money from unlimited everything subscriptions?” We’re shooting for the pathetically low goal of trying to maximize the revenue we can get from a single flawed business model. That’s the wrong question, and no matter what answer you pick, the music industry loses. 

Instead, we should be asking

The answer is by finally taking advantage of the customization made possible by the internet, and getting serious about capturing the tremendous value created by the music industry. Stop going with the easy short-term solution, and start pursuing the right long-term solution. 

—— 

*http://www.hypebot.com/hypebot/2013/01/tag-strategics-ted-cohen-on-music-tech-and-the-music-industry.html

^http://rockonomic.com/2015/01/19/wayback-machine-why-music-service-prices-are-falling-and-cant-get-back-up/

**We’re leaving out any discussion of the ad-supported free tier here; our example assumes a simple option of “unlimited everything” subscription for music, or no music. The option of the ad-supported free tier complicates the analysis significantly; however, it needs to be addressed because it’s a factor, and we will address it. But just like in math, you’ve got to get the fundamentals down before moving on to the advanced conversations. 

See website for complete article licensing information.