Image: Tesla

Nearly a year ago, Gartner predicted that “connected cars [would soon become] a major element of the Internet of Things (IoT),” with 20% of cars sporting a wireless data connection by 2020. With mounting competition fostered by Tesla, already that number seems light.

But, what Gartner didn’t catch, and what Andreas Gissler, Managing Director of Automotive Strategy at Accenture, also misses, is just how much of the data generated by cars will pad the pockets of companies like Google and Apple, instead of the automotive giants.

The reason is portability. That is, the companies that best enable our car data to speak to the rest of our digital lives will ultimately own that data.

Driving computers around

Today’s in-car data crown goes to Tesla, which is as much computer as it is car. Unlike other brands, which still treat data as an after-market add-on (e.g., my Honda Pilot requires me to order a CD annually with map updates for $150, rather than delivering it to me for free over-the-air as part of my regular service), Tesla makes data the essential attribute of its driving experience.

Early on, Tesla disciples cheered when the company used in-car data to refute a potentially damaging New York Times story. Since that time, however, it has become clear that Tesla’s use of data has little to do with public relations and everything to do with improving the driving experience.

For example, Tesla’s cars are constantly learning. While users reported early problems with its new autopilot feature, they later reported continuous improvements to the service. As Tesla founder Elon Musk recently said in a press conference, “The whole Tesla fleet operates as a network. When one car learns something, they all learn it. That is beyond what other car companies are doing.”

Well beyond.

Tesla may lead the pack when it comes to data use, but other automobile companies aren’t sitting still. Following Tesla’s example, mainstream carmakers are increasingly threading data into many more cars. As Gartner finds, “The connected car is already a reality, and in-vehicle wireless connectivity is rapidly expanding from luxury models and premium brands, to high-volume midmarket models.”

Which may not prove to be as big a boon to these automobile vendors as they’d like.

Who will own your car data?

Historically, Americans tend to buy cars every 4.6 years. But, since the 2008 recession, we’ve become more frugal, holding onto our cars for 6.4 years. In fact, according to AutoMD, a large percentage of us will keep driving the same car for upwards of 10 years.

Imagine using the same computer for 10 years. Or 6.4 years. Or even 4.6 years.

For most of us, tuned to trade in our primary computing device–our smartphone–every two years, it’s unthinkable that we’d still be using the equivalent of our Motorola Star-Tac while the rest of the world plows ahead with the iPhone. Yet, this is essentially what the carmakers expect of us.

Think about that for a minute. I have an in-car mapping system with my Honda Pilot, but even after three years it’s too cumbersome to use. Instead I turn to Google Maps to get me around, because both the software and the hardware supporting that application refresh constantly.

True, companies like Tesla make this less painful by constantly updating their software on the fly, but it’s going to be difficult for any company–even one as forward-thinking as Tesla–to keep up with the pace of consumer technology, given how infrequently we update the hardware we drive.

As such, I’d expect to see Google, Apple, and other mobile behemoths steer our automotive future. The carmakers won’t want to give up their data, but unless they can find ways to make it as portable as Apple can, what choice will they have?

One option might be for the traditional automakers to experiment with non-traditional conceptions of ownership, opting to push subscriptions (like leases) that either make ownership of a particular car transitory or completely irrelevant (think: Zipcar). But such a fundamental shift in their 100-year-old business model will be even more difficult a transition than learning to make data the center of the driving experience.