As the Chief Economist for PRS for Music and one of the few actual economists in the music business Will Page has a reputation for providing clarity, both on the state we’re in as an industry as well as the direction we should be heading. PRS for Music is one of the largest collecting societies in the world, representing some 60,000 songwriter, composer and music publisher members, collecting and paying royalties to them whenever their music is played, performed or reproduced.

Will and I actually first met over a beer at a music venue I was booking at the Edinburgh Festival in 2005, when he was working as a music journalist for Straight No Chaser. We have both taken somewhat drastic turns in our careers since then and, by happy coincidence, Will stumbled across this very blog and decided to get in touch to reminisce. We have been chatting ever since about his work, particularly with regard to it’s relevance to China.

Will Page

His latest report The Long Tail Of P2P, co-authored with Big Champagne’s Eric Garland, was presented to much fanfare at this year’s Great Escape Festival in rainy Brighton, UK, an event I was lucky enough to be invited to attend (Thanks Jon McIldowie and UKTI). Will has kindly agreed to me running a few questions by him on the subject:

Ed Peto: There has already been a good deal of coverage on your work on the demand curve for digital music consumption - from New Scientist to the Financial Times - particularly with regards to your contention of Chris Anderson’s Long Tail theory - but, for the benefit of people who haven’t read it yet, could you give us a quick elevator version of your latest Long Tail Of P2P report and its findings?

Will Page: Sure. The original Long Tail concept, as laid out by Chris Anderson in a famous Wired Magazine article in October 2004, goes like this: If you offer people more choice, and help them make that choice, they will take that choice. It proposed that in a world of widespread Internet access, it no longer makes sense to cater to the public appetite for the most popular CDs, DVDs and books. Instead, even the interests of the smallest niche might now be served. In short, the tail of available niche products would lengthen (supply-side effect) and then fatten with sales (demand-side effect). And so the “Long Tail” emerged.

To recall, Anderson’s theory relies on a change in the nature of the supply curve given barriers to entry falling and a great many new products can now get to the market. However, it takes two curves to tango in economics, and consideration of the demand curve completes the picture. What we uncovered from that analysis was a shock to some and no surprise to others: a ‘hit-heavy, skinny-tail,’ log-normal distribution for legal online music consumption; a distribution not that dissimilar from what one might expect from a more traditional, bricks & mortar store.

This dormant tail, pinhead pattern appeared across a number of digital music providers, in the markets for singles, albums, as well as streams - the three markets for legally consuming music online. But of course the illegal music market has been with us for longer, and is considered to be much larger than the legal one - so the next intuitive step was to understand the shape of demand in P2P. What we uncovered was another hit heavy skinny tail distribution, and that’s what we presented at the Great Escape. The results raised a few eyebrows, that’s for sure.

EP: Here in China we see also see an incredibly head-heavy distribution curve, with pop hits dominating the musical landscape. I tend to explain this by suggesting that, in China, music is used as a way of fitting in and not as a differentiator as it often is in the west. In short, the reason, I believe, is largely cultural (with censored media being another contributing factor).

According to your research, however, western consumers also largely seek out hits even when presented with the essentially infinite choice offered by illegal services such as P2P. Do you think that, like the Chinese, western consumers also have a deep-rooted cultural proclivity for hits, or is the behaviour you have identified in your study a hangover from a period of limited inventory, limited access and bottlenecked media and marketing? ie. Is it nature or nurture?

WP: That’s a great question. Firstly, the fact I’ve uncovered this hit heavy distribution for music does not mean the Long Tail is dead - there may be other examples of ‘fattening’ tails in books, film and television. But then perhaps that’s the point - some forms of media goods are for sharing (i.e. music at a festival) and others are for private consumption (i.e. a book on a train journey). Maybe that’s why ‘Book Clubs’ still haven’t taken off in a social networking era?

Now to your question. I don’t think it’s a hangover - there have always been niche markets, and one could argue that they were more effective prior to the long tail era kicking in. For example, I wrote for the niche music publication Straight no Chaser for seven years, and spent a large amount of time digging for rare Brazilian and African vinyl.

Point being, the magazine has closed (advertising revenues in a digital age) and lots of those niche shops have closed down. So I reject the hangover assertion, there may well be examples of the tail being a lot fatter prior to the book coming out. On that note, let me also add that you have to really think about the quality of data, both then and now. Niche music products are often purchased in second hand record stores - I can testify to that as I practically live in them! Not only is there no data on second hand sales, there’s no copyright either. That’s an important dynamic in an online physical world like Amazon, where first and second hand goods are priced side-by-side. A fat or skinny second hand niche market is therefore (i) hard to prove and (ii) even harder for artists and songwriters to benefit from. It’s an anomaly that’s really worth pondering. .

Another angle towards answering your question is to consider the tools which are being used to understand media markets like music, film and books. I mean this whole Long Tail debate has been dominated by economics, and us economists are terrible at losing sight of reality. Another angle, which we raise in the paper, is that of ‘culture’. On that note, I’m inclined to cite Andrew Bud, the Executive Chairman of mBlox, who has been like a professor to me in pioneering much of this long tail work to date::

“…It means something that we are seeing a log normal distribution in the sales data for tracks. That only happens if the more successful a track becomes, the larger are the random forces affecting its sales. But then the question is how does the market know how big a track is? Why does the scale of a track’s success matter to the choices people make? An obvious answer is that it’s through people chatting to each other and seeing the music talked about in the media. That’s what culture is. So the fact we’re seeing the log normal distribution here may point to the power of culture on people’s choices. Whereas Chris Anderson’s hypothesis of a Pareto power law would be much more about random, individual choices - people alone with their computers. So perhaps, this debate of thick versus fat is really about the power of culture in determining demand…”.

- Andrew Bud

Andrew Bud

EP: Are you able to project future behaviour from this research?

WP: No. We have not attempted any projections or forecasts based on the analysis. The original singles, albums and streaming data sets we worked on were for the twelve months from 2007 Q4 to 2008 Q3. We kept the same time period for the illegal P2P file sharing study, to keep it consistent. What we’re doing now is to look at data sets concluding in 2009 Q1 - so whilst we’re not essentially looking forward, what we can now provide our management team with is monitoring and interpretation of the changes in demand over time.

On that note, co-author Eric Garland raised two concepts in the study which merit attention here: the primacy of listening and music hoarding. To recall, these trends lead to a peculiar irony: widespread listening to music that is never stored coincident with vast stores of music to which no downloader ever listens. I think you can use our rigorous long tail analysis and these two concepts to debate future behaviour. ‘Hoarding’ especially - that’s an incredibly important concept for the music industry to get its head around.

Eric Garland

EP: Do you think that once music recommendation/discovery services have fully developed you will still see the same head-heavy results as you are seeing now?

Consider the following hypothetical online music platform:

• Every track in the world is one click away, with negligible download/buffering time.
• Each user’s music preference profile is perfectly mapped and updated continuously in real time according to their actual listening habits, as opposed to music they just download and then 'hoard’.
• Music is recommended to this user purely based upon this profile (and other users with similar profiles).

WP: The first thing to appreciate is that it could go either way - ‘good’ recommendation tools could fatten tails, or concentrate activity around heads. What’s going to be fascinating is that we’ll soon be able to answer your question with evidence. By that I mean that excellent sites like We7 and Spotify have gained incredible traction already this year, and that will allow them to further develop their offerings in line with the customer’s demands. From there, we can see what demand looks like, given the infinite choice from supply. .

There’s an important point to be made here, though - which is when critics have dismissed my work by saying that a long tail market without a good discovery tool is just noise. I mean, sure, I take the point - but I’ve got to counter it, as it implies ‘when the facts don’t fit the theory, then there has to be something wrong with the facts’. The objective, surely, is for these promising music sites to become profitable first and foremost, whereas fattening the tail is an optional extra. If the latter results from the former, cool - but it doesn’t necessarily have to work that way.

What I’ve seen so far - and by that I mean some of the staggering success stories of digital music in 2009 Q1 - suggests that the idea that ‘when you offer people more choice and help them make that choice’ their behaviour is a lot stickier, and their willingness to roam a lot more tamer, than the theory would have had us imagine. My colleagues Chris Carey and Gary Eggleton (who are both far brighter than me) think that our work in this area now has us close to helping the music industry understanding the limits of unlimited choice. That’s really exciting as we’ll be able to offer our songwriters and publishers important new insights that they wouldn’t have had otherwise.

On that note, I’d like to quote psychologist Barry Schwartz who summarizes his excellent book, The Paradox of Choice, in a recent TED lecture: “There is no question that some choice is better than none, but it does not follow that more choice is better than some. There’s some magical amount, I don’t know what it is but I’m pretty confident that we’re long since passed the point where options improve our welfare”.

EP: How do the results of this research impact upon your work at PRS for Music?

This question is two-fold: what does it mean for PRS for Music and what does it mean for its stakeholders - the rights holders and users who we bring together. I think you can see three applications of the long tail work, those being costs, segmentation and investment strategies. With regards to the latter, there are some fascinating debates to be had. For example, Anita Elberse has looked into why you get irrational bidding wars in the book publishing industry, even when the market is not in a healthy state. Her work is really inspirational and I’d strongly recommend your readers check it out. My interpretation, for the music industry, comes down to this - if you’re in a market affected by the long tail, do you bet large, bet small or do you bet at all.

One final point, though, is this. My work is not ‘anti’ long tail, nor does it have anything to do with ‘bashing Chris Anderson’ - the press love a Punch and Judy show, but this is about understanding markets. Let me reiterate, I really rate the Long Tail Book and recommend it to anyone who hasn’t yet read it. Moreover, Chris Anderson’s ‘blog’ was an excellent tool for engaging people like me into the debate that we would otherwise not have known about. I’ve always said that as soon I find real evidence of the long tail at work, Chris will be the first to know and I’ll be the first to celebrate. There’s another collaborative project we got going here in London, it’s wrapped up in confidentiality just now but the way things are beginning to look, I should be letting him know very shortly!

6Interview, Staff Blog, Ed Peto, General Music Industry, Large,