the meta-roj music store
online music stores quickly converged on 99 cents, because it’s close – real close – to making a little money. close enough that you can fudge it in the business plan. i like fudge, so, today, i’m going to launch the meta-roj music store.
the top line
based on previous material, we know that the top-line is going to converge quickly at 99 cents per track, unless i decide to take on walmart at 88 cents.
i’m going to make an executive decision that we’re going with the pack on this one, at 99 cents. that could make me either greedy or conservative. maybe both.
what does the music cost?
i’ve seen at least two credible breakdowns for online music sales:
The download services pay music labels around 79 cents per track in royalties.
billboard/reuters [pdf, via delvian records]Sources present at the meeting say the terms offered to the indies were identical to those granted to major-label participants: a 65%-35% split of proceeds from the 99-cent downloads, favoring the labels, with payments every month.
for the meta-roj music store, let’s say that i make the label’s cut 72 cents (that’s about halfway between these estimates).
what does the sale cost?
now that we have the music, we need to sell it, and that introduces transaction costs. since i’m not a credit-card company, i need to make a deal with one. typical terms from the big-two (visa and mastercard) are 25 cents plus 2.5%. (walmart has had some success taking on the big credit card companies, so, maybe there’s some negotiating room).
to really estimate this expense, we need to know how many tracks get sold in a typical transaction – and that’s going to be hard to pin down.
a one-track sale at 99 cents would cost 25 cents + 2.5 cents = 27.5 cents per track.
a two-track sale at $1.98 would cost 25 cents + 5 cents = 15 cents per track
a 14-track album sale at $9.99 would cost 25 cents + 25 cents = 3.6 cents per track.
i don’t have much leverage, and i think many sales will be single-track situations. then i need to add in some other per-track and per-sale expenses, like the fractions of pennies that add up as i transfer the few-megabytes of song data and such.
let’s say that my transaction costs come in around 18 cents per track (because that’s just going to make a nice round number).
ya gotta build it before they come
now, we come to the issue of building a big site and keeping it running. some servers, some code and some labor to get everything ripped, encoded, tagged, encrypted, rights-managed and ready for sale. and i need to promote this thing, so you’ll find it.
$500,000? $1,000,000? let’s go with a million.
the bottom line
i get 99 cents per track, but it costs me 90 cents on each one. my margin is 9 cents per song, and i’ve got a $1 million hole to dig myself out of… it takes 11.1 million tracks at 9 cents each to get to $1 million.
Between June and November, music lovers bought 7.7 million songs online, but only 4 million single-song CDs at stores.
that’s enough to pay for about 2/3rds of the meta-roj music store, in just 4 months! and the business is accelerating! only 13.3 other stores to go. itunes recently announced their 20 millionth track, but you have to dig a bit to decide how the dollars fall out from that mark. with 750,000 albums at $9.99 and 11 million singles at 99 cents, that means itunes has grossed about $18.4 million. if itunes follows our 10%-ish margin, that’s $1.84 million. now how much did apple spend to build it and promote it?
eh, no matter. those crazy kids love their music. plenty of room.
what’s the potential?
we have some data on the global market for recorded music – about $30 billion per year. that is based on 92.5% of sales as albums and only 7.5% as singles, revealing an average unit-price of $9.8.
unfortunately, the meta-roj music store is in the singles business (the evidence is hard to ignore). if it’s anything like itunes, it’s more like 5.5% albums and 94.5% singles. the itunes experience suggests an aggregate unit price – singles plus albums – is about $1.49.
from here we can make two assumptions – the $30 billion is going to stick. as they are given more choices, and more options, people will buy more music. they’ll buy just as many dollars worth of music as they did before, so we’ve still got a $30 billion pie to split up. at 99 cents per track, that’s 30.3 billion tracks sold per year, so there’s plenty of room for the meta-roj music store in the market. i only need .037% of the market to break even ($11.1m/$30.3b). sounds just like every other “if we just capture 1% of the market…” pitch!
the other possibility is that they will buy about the same amount of music as they did before. and then we get two more paths to follow, both of which lead to some more-troublesome math (and a lot more assumptions).
let’s start with unit-sales of 2,850 million albums and 230 million singles. if (and this is a great debate) each album really has only one or two good tracks, let’s give them credit for three “equivalent singles per album.” based on that assumption, we’ll call the market at 9 billion tracks (8550 + 230) per year. we’ve just watched the market collapse from about $30 billion to about $9 billion. big change, that.
another way to come at this is to just work the unit price data. $30 billion at $9.8/unit is 3 billion units. 3 billion units at $1.49/unit is $4.5 billion. even bigger change, but i can still make the “1% pitch” even at this level – $11.1m/$4.5b = .25%!
there’s room for everybody
even with the $4.5 billion market, and 14 competitors (that’s the 13 announced players plus the meta-roj music store, of course), we can split this evenly and each take $320 million! with 10% margins, we net $32 million! each! every year! and, of course, the meta-roj music store is going to get the lion’s share of the market. we should be able to double or triple that!
… and that begs the question
it’s an easy pitch. you can flesh this out in several dozen different ways to make it look good. it can look good for apple. it can look good for walmart. it can even look good for coke. that’s why 99 cents. that’s why everyone’s piling into this game. now, the important question that’s missing from this post: what could go wrong?