ad-sponsored music and the major labels

What are the economics of ad-sponsored streaming music services like iMeem? Labels want some amount, most commonly quoted as a penny a play, and the question is whether this price makes sense, or if not what the price will have to be.

A song lasts 3.5 minutes. The majors have been asking $.01 for it. The site pays for the play by selling advertising. Let’s say the site shows a new ad every time the song changes. To break even the site needs to sell one ad per song at the rate of 1 penny a song, which gives you an effective CPM (“eCPM”) of $10.

A $10 eCPM isn’t feasible. Sites don’t earn that kind of rate with 100% sell-through. Even if were feasible it leaves no room for the rest of the business. They have other costs. They need to earn a profit, and it has to yield a return on investment comparable to web businesses that don’t pay music royalties; otherwise investors will move their money out of music-related products into royalty-free products like search engines.

A $1 effective CPM is closer to the mark, meaning that Myspace, Google, Facebook, etc need a 10X price reduction from the labels to make this business work. The labels see this as unreasonable. They’re already lowering prices from what they earn at the iTunes store — why should they keep going to accomodate third party businesses at their own expense?

The alternative business for the labels to be in is selling music by the piece. The majors gross about $.70 on a download at the iTunes store, 70X what they are proposing for an ad-sponsored play; $.01 is only 1/700th of that! So naturally the price change is freaking people out.


But of course price is irrelevant by itself. Profit is a function of price, marginal costs, and volume. In the recording industry marginal cost is close to zero, and fixed costs like studio time don’t go up with volume, so the basic equation is price*volume.

Music plays which feel free because they are sponsored by advertising should get a lot of plays. If you could search for a song and play it right there, without all the friction of paying for a download or even downloading at all, you’d be more likely to go through with the transaction. Price would be lower, but there would be more transactions. And the only question that matters is whether the number of ad-sponsored transactions would be high enough to make up for the lower revenues per piece.

If the price is 1/700th, there needs to be 700X more transactions. A gold single with 500K sales would need 500K*700 = 350 million plays to get into the black if all those sales were converted to ad-sponsored plays paying a tenth of a cent per play.

What’s the likelihood of this working for the labels? Will ad-sponsored plays be at least 700X more popular than sales at the iTunes music store?

Almost nobody buys CDs; only 15% of the population; and the number of people who buy singles from iTunes is miniscule. But almost everybody listens to the radio. Ad-sponsored plays would extend the reach of monetizeable events to people who don’t buy CDs and to people who aren’t committed enough to a particular release to buy it. For the fan club members who listen to a CD over and over again, every play would be monetized. Plays in an incidental context like background music in a sports site would be a worthwhile business. Ad-sponsored plays in the browser would replace much filesharing, because they would be so much more convenient, and this is a gargantuan pool of monetizable events.

So the volume is indeed likely to be very high. Because these numbers are very fuzzy there’s no way to say whether the volume growth will be 350X, 700X, or 1400X. Still, they’ll easily get into the right range. The necessary volume isn’t guaranteed but it is realistic, and there is plenty of upside because of the possibility that volume growth will blow way past 700X.

Internet strategy

But there’s one other issue in addition to volume: compatibility with the web. The web is almost everywhere, and it needs music to be able to flow with no more friction than any other content. It can do this with or without the majors on board, and unless they choose to get on board it will leave them behind.

Sites that try to comply with label requests repel users and soon go out of business. There is no way users are going to stop for a paid download whenever there’s a need for a song. First of all, acceptable page load times are in the range of a few seconds, and when you add on the time to acquire a download the load time goes up to minutes. Monetizing via ads is the only way for the labels to support acceptable load times. Second, interrupting the user to incorporate a paid download into page rendering creates fatal usability problems. Users will get confused and leave the site. Third, there’s no way to enforce the use of paid downloads without restricting available music to what the iTunes store will carry, and this is not a web-scale design. Because these are fatal problems, successful sites route around the labels by having users of music-related features fall back to filesharing for provisioning.

Can labels counter filesharing networks without moving to an ad-oriented revenue base? It depends on how advertising compares to other revenue sources. Labels can sell complementary products like concerts and merch, but other businesses already occupy those niches. They can earn a cut of a music tax to cover filesharing, but that’s a wildly speculative idea which the labels don’t support because it would undercut their per-unit sales. They can take part in other experimental blanket licensing schemes, but they don’t support those because they undercut per-unit sales. Labels can get paid for subscription services like Rhapsody, but these are and will always be a small niche. They can earn a living on CD sales, but these won’t support a very big recording industry. They can earn a living on download sales, but these aren’t growing fast enough to replace CD sales.

The only bright spot in this picture is that online advertising is thriving. It can generate the revenues the labels need. In addition it makes technical sense, since ad-sponsored music can flow as easily as ad-sponsored text, and ad-sponsored text is the majority of the web. This model works.

I don’t mean to say that per-piece sales will go away: fans will keep buying them here and there. Providers like eMusic will continue to exist. But this business will be an adjunct to the ad business, just like all premium services on the internet. The only thing that will save the major labels is to embrace ad-sponsored plays as their primary revenue source, and that requires them to drastically lower the royalties they’re seeking.