Reflections on the tail


From paidContent yesterday:

Meanwhile, Universal Music Group’s eLabs president Larry Kenswil warned that the ”long tail” – the subject of Wired editor Chris Anderson’s book illustrating how the net leverages small but everlasting value in archive material – was not enough to make up the shortfall.

“It’s an interesting catchphrase. Many more products will sell to a larger audience, that’s true,” he said. “But the question is – what is the offset of the most popular products selling less?

“It’s obvious that shelf space on the internet means people have access to things there wasn’t storage for before.

“But the long tail means that most of that will be listened to [only] once. So, even if you have 300,000 songs being listened to once, it doesn’t make up for the fall-off of the number one this year from what the number one was last year. No album sold 4m last year.”

I feel there are 2 issues with Larry’s argument:

1. Re the comment “what is the offset of the most popular products selling less?” If one considers this from an industry perspective, I do think there will be an offset – or at least not as much of a decline in unit sales as the current trajectory would suggest – thanks to a fattening of the tail. Selling MP3s on an a la carte basis will help this. But, as suggested by the example of Amazon in Anderson’s analysis, the offset will come from 100,000s of artists, many (or likely most) of whom won’t have a label deal. My hunch is that much of these long tail sales will occur via “retail on the edges” or similar arrangements, as with the MySpace + Snocap relationship.

OTOH, if one considers the offset issue from the perspective of a particular major label, it seems unlikely that *their* sale of content further down the tail will offset the dropoff in sales of their popular products. The economics of a label (as currently configured) make a focus on the long tail untenable.

2. Re the comment “But the long tail means that most of that will be listened to [only] once.” I don’t believe (or at least didn’t realize) that the long tail necessarily contemplates frequency of consumption. Chris Anderson uses Rhapsody – an on-demand service – in his research and book to present the concept as relates to music, so I suspect that’s the basis of Larry’s comment. But I do think music fans will also purchase (as opposed to rent) music down the tail, just as they do for books (Amazon) and film (Netflix). I’d like to see an analysis of eMusic or, later, the Snocap-powered MySpace service (or even iTunes).

It’s problematic to compare incremental revenues from an on-demand service with the decline in revenues from CD sales. Based on anecdotal evidence, most people seem to use Rhapsody as an expensive music-discovery service and/or as a way to listen to music they don’t like well enough to purchase. They still acquire (legally or not) music that they really want to own. So one could consider the royalties of ~$.01 per stream from on-demand services (just as the royalties from DMCA radio-style streaming should be considered) but must also include revenues from a more appropriate comparable – the sale of music online.

And to properly assess the sale of music, both the head (at a ~$.55-.65 wholesale margin per download) and the tail (at TBD wholesale margin per download) have to be contemplated. Further, there’s little evidence to conclude that current wholesale margins result in the revenue- or profit-maximizing level of sales (which, in any event, is unlikely to be achieved until DRM restrictions are eliminated).


No Responses Yet to “Reflections on the tail”

  1. Leave a Comment

Leave a Reply

Please log in using one of these methods to post your comment: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: