Automakers: Software Development Is Hard Enough - Stop Making it Harder

It's been a tough but exciting few years for automakers. After an unglamorous moment in the national spotlight late last decade, they've pulled an impressive digital coup. The connected car is becoming a reality, as automakers have invested considerable time and effort to develop software-based systems that promise to deliver a remarkable experience to an increasingly-discerning customer base. And while there are still substantial strides to be made, the industry is already thinking ahead: electric cars; self-driving cars; cars that are more computer than nuts and bolts.

The way the automaker community has embraced software for business advantage is remarkable, and certainly an archetype that other industries, like finance or healthcare - which are in the midst of their own digital revolutions - would love to emulate.

But I live in the software world; I've seen digital transformations, and I have some advice for automakers: becoming a software company is hard enough - don't make it any harder.

A Car is a Digital Device
When you think about it in the context of connectivity, your car is really just another device. You can access your music, for example, on your laptop, your tablet, your phone - and, increasingly, through your car. Of course there are more complexities related to a car, and some very unique usage models - but in the end, consumers expect (or will expect) the same type of seamless connectivity while they're driving. And within this context, the proprietary platform strategy of most carmakers starts to unravel.

When it comes to a phone, the OS is essentially a commodity; it's what's built on top of the OS - the types of unique apps and interconnectivity - that make a device stand out. Consider Samsung's boundary-pushing innovations, like its new "Edge" information stream; or how Apple elegantly syncs experiences across devices. Platforms are tough to perfect, and the mobile industry quickly realized that once something worked, it was a better use of everyone's time to drive adoption via the application layer over the top of the OS.

Yet the auto industry is still differentiating on connected car platforms - using resources on trying to perfect what will ultimately become a commodity.

A Standard Path to Differentiation
A digital transformation is already tough enough without trying to create an industry from scratch. Developing and deploying software today is an intense and demanding task. Successful apps require agile development processes and lightning-fast releases and updates - facts that bump up against the realities of corporate bureaucracy. To manage these dichotomies, automakers have developed what amount to internal start-ups, designed to develop connected car strategies with big bets, rapid development, and the latitude to test, fail, learn and succeed. GM, for example, recently hired 8,000 developers to build out a new system supporting dealerships and franchisees.

Imagine, instead, the opportunity if the industry came together to develop a standard platform (or platforms) on top of which all car makers could develop differentiating app experiences. Instead of each company battling the same issues of design, integration and, most critically, security (consider the recent headlines around white hats hacking a Jeep and a Corvette and controlling the cars), these tasks would be addressed with ten times as many experiences and solutions.

Developing a set of standards could address the issue of regular software updates (a bigger issue with a car than a phone, which gets replaced every year or so with a new model); it would regulate the way security was considered and built into software; and it would establish clear rules for emerging problems, like how to ensure that data is wiped away when a connected car is sold as a used connected car. Of course, it doesn't address the equally prickly question of who owns the data generated through interaction with a connected car; but that's a discussion for another day, as is my prediction that we'll soon need to sign End User License Agreements when we buy a car - just like we do when we sign up for Facebook or Netflix.

We are already seeing a move to this mindset: emerging examples of standardization caused by the mobile wave, like Apple CarPlay and Android Auto, are proof of recognition that people don't want yet another inferior way to navigate or listen to music; they just want to bring these mobile experiences seamlessly into their cars.

With these development resources freed up, carmakers could instead focus on developing software that is not only differentiating, but industry evolving. Their development teams could focus on areas that auto makers can control and benefit from - things like maintenance data, safety data, car performance data, driving behavioral data - while at the same time finding ways to differentiate via the software services that they deliver, like infotainment, car maker diagnostics/safety/performance solutions, geolocation services or concierge maintenance services.

And that's ambition well spent.