As all long-time eMusic watchers are aware, eMusic’s business model has always been based on the “health club” model, i.e., the assumption that a certain percentage of customers will pay for but not use the service. In eMusic’s case that corresponds to subscribers who download fewer tracks per month than they’re paying for. The result of these unused tracks or digital “breakage” (as Digital Audio Insider refers to the phenomenon) is that the per-track payout from eMusic to labels was somewhat higher than it would be otherwise. That in turn made distribution through eMusic somewhat more attractive to labels than the nominal per-track pricing might otherwise indicate.

However with the recent price increases and the introduction of “album pricing” I suspect that eMusic is consciously moving away from reliance on digital breakage. One problem of eMusic’s model from the labels’ point of view is that the per-track payouts were in no way guaranteed: If a greater percentage of eMusic subscribers happened to use their full monthly quota of downloads then per-track payouts would inevitably decline. The vast majority of the reported disputes between eMusic and labels, including those labels who’ve left eMusic, revolved around this issue.

I’ve previously criticized labels for the single-minded focus on per-track revenue and profit as opposed to total revenue and profits, but labels have at least two reasons for their position:

  • While per-track payouts might fluctuate, labels typically have fixed costs they need to cover per track, most notably for mechanical royalties paid to songwriters and music publishers. If the per-track payout becomes too low then labels lose money on every track distributed through eMusic, with no way to “make it up on volume.”
  • As CD sales continue to decline labels are under pressure to maintain their revenue and profits. All other things being equal, the simplest way for labels to survive, at least in the short term, is to try to hold the line on pricing of digital tracks and charge as much as the market will bear. In theory they may be giving up some revenue and profits by pricing tracks too high and thus forgoing sales to more price-sensitive buyers, but (at least based on a recent Billboard analysis) it may be that in practice higher prices do not hurt sales enough to offset the benefit of higher per-track payouts.

One obvious way to increase per-track payouts is for eMusic to increase the price of subscription plans and/or reduce the number of monthly downloads included with a plan. This is essentially what eMusic did with its recent price increases; the eMusic Basic plan had a reduction in number of downloads, while the Plus and Premium plans had a price increase as well. These changes had the effect of fixing a minimum per-track price of about $0.40. However the per-track payout would still vary based on the behavior of eMusic users and the amount of digital breakage in a given month.

Enter album pricing, which as implemented by eMusic in cooperation with Sony and other labels can be thought of as a way to achieve the same effect as breakage while eliminating (or at least minimizing) the element of chance associated with breakage. For example, consider a user on the Basic plan who has a monthly download quote of 24 tracks. In the absence of album pricing the user might in a given month download an album of 4 (long) tracks and another album of 8 tracks, and then not use the remaining 12 downloads. The breakage is thus 50% of the user’s quota, and the effective per-track price is about $1.00 per track ($11.99 divided by 12). However in another month the user might download 18 tracks, corresponding to 25% breakage and an effective per-track price of $0.67, and in a third month might download all 24 tracks in the monthly quota, resulting in a per-track price of only $0.50 and no breakage.

However under album pricing purchasing many albums with less than 12 tracks actually requires 12 “download credits,” as noted by many eMusic messageboard posters and discussed in a Billboard analysis of eMusic’s recent changes. In our example, instead of using 12 downloads of 24 to download two albums of 4 and 8 tracks respectively, the user could well have to spend 12 credits per album, thereby using up the entire quota of 24. Strictly speaking there was no breakage (i.e., the user spent all the credits to which they were entitled) but the resulting per-track price of $1.00 is the same as under the old plans with 50% breakage.

As more and more albums move to album pricing, it becomes increasingly difficult for a user on a Basic plan to buy more than two albums per month, and if they use eMusic at all it’s like they’ll buy at least two, and thus have no breakage in the traditional sense. In effect album pricing allows labels to remove the element of chance involved in user behavior and manipulate the per-track price themselves (i.e., by designating a particular album of less than 12 tracks as requiring 12 credits, and designating particular tracks as album-only) according to their business objectives.

It’s correct that the per-track price is effectively reduced when an album with more than 12 tracks is sold for 12 credits under album pricing. However in practice this most likely doesn’t matter, since per-track costs such as mechanical royalties typically are not any higher for albums with more than 12 tracks than they are for 12-track albums. (For the full and gory details see the “maximum rate per album” discussion in chapter 16 of Donald Passman’s All You Need to Know About the Music Business.)

So the labels win in the case where albums have fewer than 12 tracks, and don’t lose in the case where they have more. eMusic keeps the labels happy, and if some users don’t download even the few albums they can now afford (and thus help raise the average per-track price even higher) then it’s just icing on the cake.