18 Million over 3 months is nickel and diming (based on Spotify cost) ? Even for Apple it's a lot and will affect the earnings report.
Here's the math:
Let X = number of paying subscribers likely to retain the service beyond the trial. Pick your number for X. Hints: Spotify apparently has 20 Million paying subscribers now. Will Apple do better or worse than Spotify after 3 months?
We already know that AFTER the trial, Apple's cut will be about 28%. On a $10/month service, that's $2.80/month per paying user.
Let R = X times $2.80. R will be how much new revenue Apple makes per month if our guess at X is right, none of which is contributing to the earnings report now.
Compare R to that $18 Million number. Depending on what you chose for X, is R after ONE month greater or less than the $18 million?
-If R > $18 Million and $18 Million is the actual cost, the earnings report will show a positive in spite of the trial
after only 1 month of normal operations. In subsequent months, there is no $18 Million to recoup again. And this service is meant to carry on for up to indefinitely.
-If R < $18 Million, roll with another month and compare R times 2 (months) to $18 Million. If that (R * 2) is still less than $18 Million, add another month (R * 3) and repeat. If you have to go much further than about 3 months, why in the world do you think Apple is messing around with this very low revenue business? That would be so unlike Apple. In other words, increase your guess for X or proclaim Apple a fool for getting into a pitifully low-revenue business.
The point is that unless we assume relatively miserable failure by Apple to woo paying subscribers, 28% right off the top will easily recoup a number like $18 million with just the passage of (a little) time.
Another view of the same basic question: how many do we need in X to recoup $18 million in ONE single month?
$18 million / $2.80 = approx. 6.4 million paying subscribers. To fail to accumulate less than 1/3rd of what Spotify has in paying subscribers now seems highly unlikely. This is Apple we're talking about. Note that one great advantage Apple will have over Spotify is that Apple plans to embed Music as the default in iOS9. So, much like Internet Explorer killed Netscape when Microsoft embedded it in Windows, I'd think Music will crush Spotify.
My own guess for X? 25 million, just 5 million more than Spotify has now. If I guess right, 25 million times $2.80 per month = new monthly revenue of approx. $70 Million PER MONTH, or new annual revenues of $840 Million PER YEAR. That doesn't include the added hardware sales lift that such an offering is supposed to add to the sales of iDevices & Macs, which should be
something of size. Isn't greater hardware sales supposed to be what this kind of software thing is actually about?
Now, my revenue guess or your revenue guess is not profit, and there will certainly be some costs that can be directly assigned to this service. But we all know Apple wouldn't be messing around with this if their forecasted profit margin was not juicy. If we sling the R&D & infrastructure argument to try to marginalize Apple's take right off the top such that it sounds like only $18 Million could adversely affect their earnings statement, why is Apple bothering with this service?