These days, it’s easy to forget that one of Apple’s first forays into smartphones involved a 2004 partnership with Motorola to bring iTunes to certain Moto phones.

That arrangement didn’t last – and remains unimaginable today – because it runs counter to Apple’s core philosophy, which is to own the hardware and the software experience. That’s what enables Apple to command industry-topping prices and margins.

That’s also why Apple ought to be serious about building a car (with one significant caveat – see below).

Let’s imagine you had to persuade Tim Cook to fork over a billion or two to hire 5,000 thousand people and launch Apple Motors. Here’s how you could make the case:

1. Alignment with core strategy: Apple CarPlay is a half-measure, like handing iTunes over to Motorola. Getting this right will require owning the whole car.

2. “Curb appeal” is dying: “Curb appeal” is a term that is so entrenched in our vocabulary because for most of the auto industry’s 100 year history, cars moved off the lot because of exterior styling, or how they looked when parked at the curb. Have you noticed how homogenized most car designs have become? That’s largely because of safety and fuel economy regulations. The inside of the car remains much more open to differentiation and opportunity.

3. An Apple Motors dashboard/UI would propel profit: Apple has had nearly eight years to build a market devoted to its iOS software. It has sold hundreds of millions of devices running its software, devices that typically are the most expensive in their respective categories. That means many of its consumers are wealthy. Build them an iOS-based car dashboard and lots of them would show interest. I don’t think Apple could match BMW’s 1-2 million annual vehicle sales, at least not right away, but I think it could blow away Tesla, which produced 35,000 Model S units in 2014.

4. The auto industry is ripe for disruption: Lance Ulanoff of Mashable looks at the car business and doesn’t see brokenness. I worked for a few years in the automotive industry. A lot needs fixing, a view I sum up this way: our great-grandparents would recognize the business today. Car shopping can be a pressure cooker. Tenuous if not adversarial collaboration between unions and management still grapple over production. Dealers still pile cars all over their lots in the hopes of getting customers to buy today. What’s not to fix?

5. Excess production capacity = opportunity: Apple would want total control over building the human machine interface, or HMI, because user experience is its strength and the way it commands much higher-than-average hardware margins. With so many automotive factories operating with excess capacity, Apple could outsource just about every other aspect of a car’s components to the best in breed. Engines and transmissions? Buy them from Honda. Body design? Pininfarina looks awfully sharp. Leather seats? Ask Coach.

Apple Motors wouldn’t be an easy enterprise to run. State, federal and international governments impose complex, sometimes onerous demands on their production, safety, efficiency, and sales and distribution. In most states, for example, Apple would not be able to sell cars from its own stores.

Here’s the caveat I alluded to earlier. If Apple insists on building an electric, self-driving car, which would force it to sell at much higher prices than a vehicle with gas-burning engine technology, it would be a big mistake. Given how much the US government props up electric cars with subsidies, subsidies that may disappear as soon as a Republican occupies the White House, I think this would make an Apple car DOA.

If it avoids this, though, an iCar could become a profitable niche. Apple can afford to be patient. Assigning a few hundred people to wrestle with the challenges and explore solutions while it figures out the sort of car and automotive ecosystem it wants to build makes a sense for a company with over $170 billion in cash.

You probably won’t be able to buy an iCar in the next five years. Ask me again in 10 years, though, and I bet I’ll give you a different answer.