Been watching this software shift in the auto industry pretty closely, and honestly it's becoming the biggest strategic lever these legacy automakers have right now. The vehicle sales game is just too cyclical to build a real business on anymore. But subscriptions? That's where the real money is starting to flow.

Let me break down what GM, Ford and Tesla are actually doing here because the numbers are getting interesting.

General Motors went all in on OnStar and it's actually working. They hit 12 million subscribers last year with over 120k people on Super Cruise alone. That's roughly 80% growth year over year, which is solid. Fleet subscriptions doubled to 2 million, and here's the thing most people miss - that's double any competitor. Scale matters. The company is projecting about 400 million in new software revenue this year, and their deferred revenue is sitting around 7.5 billion, up 40%. That's the kind of predictable cash flow that investors actually value because it's not tied to whether people buy cars or not.

Super Cruise is their main lever for expansion too. They're pushing it across North America and planning launches in South Korea, Middle East and Europe. Then there's the 2028 software-defined vehicle architecture coming - basically a centralized computing platform that'll handle everything from powertrain to infotainment. Ten times more over-the-air capacity. That's the kind of infrastructure play that positions software as core to their whole strategy going forward.

Ford took a different angle. They're targeting commercial customers through Ford Pro, which is actually pretty smart. Fleet operators care about downtime and costs, not just having cool tech. So Ford built telematics, fleet management, EV charging, maintenance services all integrated. Their paid subscriptions grew 30% in 2025 and they crossed 1.3 million total subscriptions, up 53% year over year. Software margins are above 50% which is where you want to be. These subscriptions now account for 19% of Ford Pro's EBIT. Still small relative to total earnings but the trajectory is clear - this is becoming a real lever for margin improvement.

Tesla's doing something different entirely. They just killed the 8,000 dollar one-time FSD purchase option and went full subscription. This is interesting because it actually makes sense economically - you'd need years of use to break even on that purchase price anyway. But there's more going on. Musk's compensation package is heavily tied to reaching 10 million active FSD subscriptions over the next decade. So the move to subscriptions-only isn't just business strategy, it's aligned with how his incentives work.

The numbers are moving fast too. FSD subscriptions more than doubled sequentially in 2025. Users have logged over 8 billion cumulative miles. In South Korea alone, customers drove a million kilometers in just one month. Europe and China approvals are still pending but they're already doing ride-along experiences across multiple countries.

What's really happening here is all three are using software as a major lever to shift from one-time sales revenue to recurring predictable streams. That changes the entire valuation story. Investors value recurring revenue way more than cyclical sales because it's stable, it's predictable, and it compounds over time. The auto industry is finally figuring out what software companies figured out years ago. This shift is going to be huge for margins and earnings visibility going forward.