I just finished a driving trip to go visit family in southern California, and the time on the road gave me the ability to unplug and think things through. I’ve been a very strong proponent of the need for monitor-less desktops in the upcoming rumored refresh coming — I’m thinking March.

The rumor mill is very quiet about any hints on these new machines, but some chatter has surfaced recently suggesting what Apple will do is a higher end machine being dubbed the “iMac Pro”. What that means, exactly, we don’t know yet.

That wouldn’t be my first choice, but the more I look at the realities of the Macintosh market today, the more it seems to be the option that would make sense for Apple while still solving the needs of many or most of the users waiting for these refreshes.

Why do I think that? Economics.

I’ve heard from a couple of sources — nobody on the inside, but others who would have some clue about this — just how dominating the laptop has become within the Macintosh product line; it seems to be that as many as 90% of Macs sold now are laptops. On top of that, within the desktop line, the iMac dominates sales, again as high as 90% of units.

Thanks to Jason Snell here are some numbers. Macintosh revenue in fiscal year 2016 was about $24b, selling around 19.5 million devices. That puts the average selling price at around $1250. Overall Macintosh is roughly 12% of Apple revenues. Since about $5b in revenue would qualify for the Fortune 500, the Macintosh part of Apple is on its own one of the larger companies in American.

But when you start breaking down the numbers….

Let’s assume that updated desktops boost the percentages, and desktop sales rebound from the rumored 10-15% of units to 20%. That would imply that units sold would be around 15.5 million laptops and around 4 million desktops.

We will also assume that the percentage of non-iMac devices grows as well, up from the tiny 10-15% I’ve been told back to 20%. That means that for desktops, roughly 3.2 million units are iMacs, and 800,000 no-monitor units. Even if we assume a large percentage of those devices are higher end units and not the cheap Mac Mini units, which would allow us to assign a higher average selling price (and higher margins) to these machines, the number and revenue is small. Even making average selling price $3,000 we’re hard-pressed to turn the non-iMac desktop part of the business into more than about $2 billion of revenue a year, which sounds like a large number until you realize it’s only about 8% of Mac revenue — actually less, because we’re assuming unit and revenue improvements with the refresh that would shrink the percentage — and when you look at it as a percentage of total Apple revenue… Desktop fans would just get depressed.

The problem becomes fairly obvious: if Apple were to build a desktop line other than the iMac, it’s going to be a moderately expensive endeavor. You have to design the electronics, the cases, the packaging — there are aspects you can leverage out of the iMac but for the most part, creating a desktop unit is starting from scratch for most of the product.

Can Apple do that and make the economics work? Does it make sense for them to create these new desktops, and could it remotely make financial sense to do so? The more I look at the numbers, the more this seems to me to be a “lose money on every one, but make it up in volume” problem, except somewhat fortunately, the volume numbers are going to be relatively low.

That’s why when I started hearing about the “iMac Pro” it got me thinking. If you build the new high end desktop as a configuration change to the iMac, much of the engineering and manufacturing costs go away, because you’re leveraging a lot of the existing infrastructure Apple already has. It costs a lot less to create a “Developer iMac” than it does to create a “Developer Desktop”. And that means it’s a lot easier to justify the economics of creating these high end configurations for what is really a tiny niche.

That leads me to believe the rumors of the iMac Pro are true, because as big a fan as I am about supporting the traditional no-monitor desktop, it’s hard for me to see how those no-monitor devices make sense to put energy into. Because of that I’m coming on board with the idea of the iMac Pro as the right way to solve the need of a high end computer for most people. It’s not a perfect solution — especially if you’re looking for a Mac to stuff in a data center, say — but I think for most people an iMac Pro configuration with decent components will solve their needs if not their wants.

Now, I expect to get yelled at on a few points, so let me try to answer them up front.

You’re talking about $2 billion dollars like it’s a tiny bit of money!

Well, yes, I am, because at the kinds of scale Apple as a company works at, it is. And if you’re a business, and you are trying to decide if you focus your time an attention on this $20 billion thing or that $2 billion thing, well, the answer is probably obvious.

Apple has all that money in the bank! they can afford it!

Which is true, but the reason Apple has all that money in the bank is that they don’t spend it on things that lose money too often. Beyond that, the key chokepoint for Apple isn’t dollars, but talent. There is a finite talent pool Apple has access to, and it has to choose how to allocate that pool most effectively. It doesn’t matter how much money is in the bank if you can’t hire, train, house and enable qualified people, or if you can only hire substandard talent because that’s all that’s available.

Apple should split the Mac into its own business

On a simple basis, I don’t disagree, but the devil is in the details. Huge parts of MacOS and IOS and the apps we use on them share a common code base now. That actually is good for the Mac because it lets Mac development leverage the development going on in IOS. But it means you can’t exactly just clone MacOS into its own Git area and change the locks on the doors so some of the buildings in Cupertino. Much of what we value on both IOS and Mac today works because it’s all tightly integrated and shares a lot of common foundations, and you can’t split that apart without screwing it up massively.

Besides, even if the Mac area was its own company, desktops would still be a tiny and shrinking part of its business, so the base economics for building a desktop device still basically suck.

And the thing is if you look at the PC/Windows market, it’s headed in the same directions: Desktop devices are increasingly either server units, custom gamer PCs, or not really there: the market has moved to the laptop.

You’re just making those numbers up.

Yeah, sort of, based on things I’ve heard that I believe are fairly accurate. But I think the assumptions I’m making are reasonable given the situation and what I could tell from looking at the Windows/PC markets for guidance, and I think they put us in the ballpark. If you have better numbers, please share them. And if you don’t like my numbers, make up your own, and write about why you think they’re justifiable. I’d love to see it..