This post originally appeared on Forbes
Tesla's revolutionary recipe has been called automotive's "iPhone moment," combining batteries, software and autonomous driving into a consumer-friendly package that has a lot in common with the smartphone.
Cars are radically simplifying their powertrains with batteries and are increasingly being defined by their electronic interfaces and software capabilities. But while the smartphone parallels are real, some auto companies seem to have taken the analogy a bit too literally.
Toyota, General Motors, Volkswagen and Mercedes-Benz are all developing their own operating systems to match Tesla. Earlier this year, Mercedes-Benz previewed MB.OS, a software hub and app ecosystem that mimics the App Store, featuring in-car versions of Google Maps, YouTube, Zoom and TikTok.
The problem is, these are all things you can already get on your phone. With CarPlay available in 98% of cars sold in America already, consumers can easily extend their phone’s capabilities to the screens in their cars.
Car companies are trying to solve a problem that consumers do not have (accessing apps) with a solution they do not want (a second-rate app experience). Anyone who has used a smart TV knows just how bad it is to use a phone app on something other than your phone. In short, putting today’s App Store on wheels will not deliver the software-defined car of the future.
The software opportunity in automotive is in delivering new capabilities that are unique to the car, solving real problems drivers experience every day.
Autonomy is the biggest single change in cars since their invention.
One of the biggest costs associated with cars is drivers. In America, 230 million drivers spend about 366 hours behind the wheel a year. Putting safety issues and the misery of traffic aside, the value of a driver’s time "costs" us more than $2.2 trillion annually.
I believe autonomy will dramatically change that math, as attention-free capabilities will give drivers significant portions of their time back to them to do what they please. This is the biggest single change to cars since their invention and one that many consumers want and are willing to pay for.
According to McKinsey, a quarter of consumers want advanced autonomous driving in their next vehicle, and "two-thirds of these highly interested customers would pay a one-time fee of $10,000 or an equivalent subscription rate for an L4 highway pilot." All in all, McKinsey predicts that autonomous driving could create more than $300 billion in revenue for the auto industry by 2035.
At the heart of these lofty projections is a new recurring revenue opportunity. Autonomy is one of the few features that delivers significant value and will constantly add new capabilities over time as the underlying technology improves. As no one wants to be driven by last year’s AI, the continuously improving nature of autonomy is appealing for consumers to pay an ongoing subscription, unlike static products like heated seats.
Perhaps most importantly, autonomy is an application that is native to the car, not the phone. This is one of the few transformational apps that car companies can really own, avoiding competing with every other app developer on the planet.
Software changes the economics of car ownership through "a new car every day."
Like most machines, cars depreciate. While reliability has significantly improved over the past 100 years, this is still both an expensive and inefficient burden for car owners.
New software applications can improve a car with over-the-air software updates that add new capabilities and fix issues remotely without going back to the dealer.
While basic OTA is already table stakes, there are still major differentiation opportunities in the breadth, depth and frequency of software changes. As software can address more components of the car, the frequency of improvements increases, leading to a greater impact on the driver experience and a larger overall opportunity.
Software is hard. Change is even harder.
Autonomous driving and the software-defined vehicle that gets better every day are not particularly controversial topics in the auto industry—most people share the same vision for the car of the future. The challenge is actually delivering on that vision.
The auto industry has dipped its toe into the software world by replicating mobile apps for good reason—it’s easier. These apps have already been done before, and surface-level infotainment changes keep much of the existing car architecture intact. The bigger software opportunities are much harder to implement.
• Autonomy is one of the hardest software problems of our time, even for the most advanced software companies in the world like Google. No one has yet quite cracked a formula that delivers both safety and scale, even though a few are getting close.
• The software-defined car is a significant departure from the product car companies have perfected over the past 100 years. Car manufacturers have optimized complex machines with even more complex supply chains and just-in-time manufacturing practices. Software and electrification throw a lot of that expertise out of the window, demanding a new and different set of core competencies. Very few companies are able to make this type of wholesale transition.
• Iterative software development processes with continuous deployments are far outside the safety, regulatory and economic processes deeply entrenched in the automotive business. Everything from product development to safety certification and the dealership model is predicated on the one-time sale of a product that does not change. Regular OTA software updates upend almost every single part of how cars are built, made, sold and regulated.
The software-defined car is a seminal moment for the automotive industry, giving automakers the opportunity to transform from hardware to software businesses, with implications on revenue, profitability and long-term valuations. While challenging, those who make it through the transition will be rewarded with successful, defensible businesses in one of the world’s largest markets.