Netflix and Hulu show how Tesla could quickly build an App Store

One of the most interesting Tesla software updates that rolled out in 2019 was support for Netflix, Hulu, and YouTube in Tesla’s V10 update. When people saw these third party user interfaces running in full screen on their Tesla, a lot of people started imagining what other apps they’d like to have in their car.

Apps on a car are a tough problem. You don’t want a bunch of junk running and slowing down the computer –– that would be a security issue, among other things. You don’t want some rogue process to hack or crash your car.

The other problem is building and maintaining SDKs, and having developers that care enough to learn about the SDK. For developers to really care about a platform there has to be a lot of users, and Tesla doesn’t currently have enough to justify investing a lot in developing specifically for their cars.

But if you look at how Netflix, YouTube, and Hulu work on the Tesla it’s clear that the companies had minimal if any involvement with bringing their services to the Tesla platform. It appears that the existing Chrome browser is simply run in full screen while the car is parked.

This web standards based solution could actually be a great way to build a simple Tesla App Store. The web based “sweet solution” would solve the two problems mentioned above:

  1. First, running the app inside a browser creates a sandbox, and ensures the car is secure. Apps can only access information specifically provided for third party apps with JavaScript APIs
  2. Secondly, using web technologies means you don’t have to invest a ton of time in learning a new platform. You can use existing skills and your existing codebase, so it’s easy to justify creating a simple in car experience

Certain functionality could be implemented as custom JavaScript APIs that are injected into the Chrome JavaScript engine to give apps some information and control over the car. For example, you can imagine an API that provides battery status, allows you to access navigation information, set a new destination, make a phone call, etc.

Tesla could curate a little “App Store”, which would basically just be a directory of URLs and some meta data about icons, etc. Some portion of the HTML, CSS, and JavaScript could be cached locally for offline access and faster loading. Tesla could approve and feature apps made by developers and customers.

Many software developers drive Tesla cars because they like how the car is made of software (as much as possible). If it were possible to build your own full screen web apps with a few affordances in the car OS to make them feel more like “apps” (similar to the Tesla entertainment section), many hobbyists would build many interesting things. Actually, many people already have developed websites with tools specifically for Tesla cars. I see a lot of blank spaces in that Entertainment screen just begging to be filled. And I have some apps I’d love to build to fill them.

Why would Tesla want to do this?

Question on Q3 Earnings Call about Plans to Monetize Tesla Software

On the Q3 earnings call, an analyst asked about Tesla’s plans to monetize its software platform. Elon replied saying their goal was just to make the Tesla entertainment system the most fun you could possibly have in a car.

More apps like Hulu, Netflix, and YouTube would be a lot of fun. Streaming services like HBO, Disney, and others seem like a no brainer… but obviously people will want to rearrange and highlight the ones they like and subscribe to. There may even be an opportunity to monetize access to those in car eyeballs similar to the way search engine makers will pay browser makers to be the default search engine.

As Tesla vehicles become increasingly autonomous, there will be a huge opportunity for Tesla to own the digital user experience for a good chunk of the consumer’s day. People spend a lot of time in their cars.

Tesla needs to be charging full speed ahead on this. In the way they saw five years ahead designing the Model 3 and designing their own FSD chip, Tesla needs to now look five years ahead and start a software platform development effort of the same magnitude as iOS saw from 2007 – 2017. The autonomous transport revolution is at least as significant (if not more) than the mobile phone revolution of the last decade, the world wide web revolution of the decade before that, or the personal computer revolution of the decade before that.

There’s a lot more to do, because people are generally spending a couple hours a day on average in the car. That’s a pretty high percentage of their waking time. […] It’s a lot of time. I guess there’s some way to monetize it, but we haven’t really thought about it that way [yet]. […] We’ll see where that leads.

Elon Musk

Based on the answer, it almost sounds like they haven’t been thinking about this enough. I’m sure some people inside the company have. Maybe they don’t want to drop any hints. But whether they’ve been thinking about it or not, Tesla is just a few short steps away from building the next hot software platform.

Personally, I can’t wait to see what everyone builds. Even in between the Tesla updates, there will be tons of new things to play with from third parties. In other words, it’s the most fun you’ve ever had in a car. And when investors start to see the potential of this software platform in an autonomous future, it’ll be the most fun they’ve ever had.

Today Tesla’s business is about selling cars. Tomorrow the business will be about selling services –– many of them primarily software based.

Software is eating the world. And now it’s eating the car.

6 thoughts on “Netflix and Hulu show how Tesla could quickly build an App Store

  1. A web-based app platform is not a new idea at all: from the top of my head, Palm’s WebOS, Samsung’s Tizen, and Mozilla’s FirefoxOS all attempted this in the past… So in a way it’s a proven concept: but also one that thus far never managed to gain a foothold — though that’s probably not really related to the technology…

    1. totally, but the palm webOS thing was a much more complicated scenario. Performance was prohibitive on a mobile device back then. I could point to electron, etc as counter evidence that web-as-native platforms can sometimes work out (although I hate electron apps)

  2. The real reason Tesla needs to introduce an app platform has nothing to do with monetisation possibilities for either Tesla or app developers. Consequently, the size of user base doesn’t really matter all that much either. It’s simply about giving users the ability to create additional functionality, and making it available to others.

    People love their Tesla’s — and a vastly oversized portion of them are software developers. And whenever software developers love something, they will create software for it, without expecting any kind of monetary compensation. That’s already happening with companion apps — but in-car apps would bring this to an entirely new level. It would give Tesla cars a further edge over the competition.

    It makes me somewhat nervous that for all his brilliance, Elon seems to be entirely missing the importance of this 🙁

    And it needs to happen soon: with Android making it to cars starting next year, they will begin building an ecosystem of their own. If Tesla moves fast, they could have an advantage that would be tough even for Android to crack (just like it was hard to crack Apple’s AppStore early advantage) — but if they keep sitting on their hands, and only try introducing a proprietary Tesla App platform after Android is already established, it will forever remain an irrelevant niche platform, much like Blackberry or Windows Phone…

    1. I don’t think they’re missing it — they seem aware but are doing it carefully.

      It would be great to have a true native SDK but a web standards based platform would be a great way to get started

      1. Well, I sure hope that’s true… However, Elon’s answer when asked about it in an interview a couple of months ago (don’t remember which one it was), sounded like he only thinks of this as offering a platform for commercial providers — which would require a much larger user base to become attractive — and doesn’t consider the potential of volunteer developers at all 🙁

Leave a Reply