Why Music Service Prices Are Falling And Can’t Get Back Up
Guest Post by David Touve on Rockonomic
Spotify recently announced another solid uptick in subscriber numbers, growing from 12.5 million subscribers back in November of 2014 to 15 million as of January, 2015. In response, folks like Mark Mulligan, Jon Maples, and others have continued pondering the impact of price (e.g., $0.99/month promotions, Family plans, etc.) upon demand for music service. I thought I would startup the Somewhat Wayback Machine and try to add further to this conversation about music service pricing using some data and analysis from a few years back…
This post is based upon a presentation I made in May of 2011 (video) at the World Copyright Summit, at which I was asked to speak on the value of music. During that presentation I offered up an analysis of data, these data suggesting that $9.99/month (or £9.99, or €9.99) was most likely not the most revenue maximizing price for music services, such as Spotify, Rdio, Rhapsody, etc. This work also led to a brief post in 2012.
Essentially, I argued that prices for these services would fall, or at least should fall, given it appeared (from the data) that the largest pools of revenue—and therefore royalties—could be found at lower prices. In other words, it wasn’t that no one would pay $9.99 for music, but that so many more people would be willing to pay at lower prices.
Most importantly, the total value of the revenue pool (i.e., price multiplied by the number of subscribers) seemed to find its maximum at prices below $4.99/month, if not as low as $3.99/month. In fact, these data suggested that by cutting prices by more than 50% the industry could make greater than four-times the revenue—the sort of circumstance that suggested significant price elasticity (sorry for the jargon).
In other words, maximizing the total value of music could mean significantly lowering prices. In a later post, we can chat about the forces that make it difficult for the music industry to think solely about total revenue.
However, as music’s largest stakeholders were contemplating the global expansion of licensed music services, a presentation suggesting the price of these services would likely fall was not exactly what everyone wanted to hear. Spotify (launched in Sweden in 2008) was just about to officially launch in the US in July, 2011, two years later. Rdio had launched about a year earlier, in 2010. Deezer had been operating in France since 2007. But in the US, we had seen music services like Rhapsody and Musicmatch (remember the Yahoo! music service?) since 2001—and prices had been consistently falling over time.
And, since I always show my work, here goes…
The data and the analysis, in pictures
While I had done other pricing work using a range of underlying data, this presentation employed only public data as published by NewMediaAge in May of 2011 (not available online). Below are the “bins” into which answers fell on a survey to which people were asked to respond to a simple question: “How much would you pay for a subscription music service?”
According to these data, about 1% of respondents would pay between £11 and £15.99 per month for a subscription music service, while about 5% reported being willing to pay between £8 and £10.99, and so on. Forcing answers into bins isn’t entirely ideal, and answers to surveys are less reliable than real behavior, but these numbers get us headed in a direction.
This direction is important if we are trying to understand whether extreme pops in subscriber numbers related to extreme drops in short-term pricing (e.g., Spotify’s $0.99/month new signup offer) are entirely a function of consumer response to “on sale now!” promotional plugs or also a signal of where demand might be at lower prices for these services. Remember, if consumers do do math, then paying only $0.99 per month for three months plus $9.99 for the remaining nine months of the year reduces the effective cost per month on an annual basis for the music service experience to roughly $7.74/month.
Using these data, we could make a simple representation of demand (see below):
To get the chart above we simply plot a point for each bin at a price at which we “clear the market” for all prices above. For example, at a price of £7.99 just about anyone willing to pay £8 or higher would become a willing buyer, and so we get to sum both bins—5% (those in the £8-10.99 bin) and 1% (for those in the £11-£15.99 bin)—leading to about 6% of the respondents being willing to pay at £7.99 per month.
At this point, we can simply multiply the price point by the estimate of the number of willing subscribers to get the total pool of revenue possibly available at each price. Since these were data from a sample of UK citizens, I used an estimate of the UK population at around 63,200,000 people from the Office of National Statistics. We could also modify this estimate assuming only those above a certain age, or only internet connected households were in the mix, or only mobile phone subscribers, or only smartphone subscribers, etc.
To be clear, what is important here isn’t the precise numbers at each price point, but rather the shape and elasticity of demand — to what extent the number of subscribers might shift given changes in price. Just a quick scan of the movement in Total Revenue for each price point suggested rather impressive elasticity of price — some % changes in price seemed to earn quite significant and positive % changes quantity (in this case, subscribers) and, therefore, total revenue.
[If you have to use voodoo to believe these numbers, here’s some point estimates extrapolating on the line (rather than estimating a curve): At 63.2 million UK citizens, we would guess 1.45 million music subscribers. Refining a bit… Using 53.5 million UK citizens above age 15 as the population anchor, these data anticipate around 1.2 million subscribers at the £9.99 price point. Shift the anchor to the RAJAR estimate of the radio listening audience (47.6 million), and you get just over 1 million subscribers. Feel free to use Spotify UK’s revenue numbers to test these estimates.]
Moving on, let’s highlight anticipated upticks in subscriber numbers based upon price movement:
Using only the points on the line (i.e., without extrapolating quantity demanded at other price points), these data suggested that the £3.99 price point would derive nearly ten-times the revenue of a £10.99 price point. That’s elasticity—a 50% drop in price links with a 300% increase in quantity demanded.
If we base our comparison against the estimate of subscribers at the £9.99/mo price point (1.45 million subscribers), revenue grows from £14,521,464/month to £57,172,710/month — a pile of revenue that is four-times larger than estimated at the current price point.
It appears (from the data) that the largest pools of music service revenue—and therefore music service royalties—could be found at dramatically lower prices.
Most importantly, the total value of the revenue pool (i.e., price multiplied by the number of subscribers) seems to find its maximum at prices below $4.99/month, if not as low as $3.99/month.
In other words, maximizing the total value of music could mean significantly lowering prices.
In a later post, we can chat about the forces that make it difficult for the music industry to think solely about total revenue.