FacePhone, the Facebook Phone

Emre Sokullu
Feb 6 · 6 min read

Facebook and Apple are officially in a war . After years of jabs between the two companies, Mark Zuckerberg finally pointed out loud and clear in Facebook’s last quarterly call that they increasingly see Apple as one of their biggest competitors. That’s mainly because of iMessage, which is bigger than Messenger and WhatsApp in the most profitable markets, but it is also because Apple’s new privacy policy on targeted ads will hurt $FB terribly. After all, the social media giant heavily relies on data-hungry advertising, much more so than the rest of the Big Five:

This chart is from five years ago. While the four on the list have a far more diverse set of revenue sources today, Facebook still depends heavily on advertising only.

While Microsoft, Apple, and Amazon are healthily divested, Alphabet and Facebook still seem vulnerable, each with a single major revenue source of advertising. However, Facebook is far more vulnerable, with

(a) Commerce (Instagram Shopping and Facebook Classifieds) still to pan out,

(b) negligible SaaS (Facebook Workplace), and

(c) long-bets like Oculus (VR, a new computing platform) and Libra (blockchain/payments) being, well, long-bets.

In this article, I will dig into the potential of Facebook entering the mobile-OS space with a smartphone of its own. This may sound like Mission Impossible because of so many past failed attempts:

  • Microsoft’s failure to make a dent as a late entrant with its decades-long desktop champion Windows, despite its mega-acquisition of Nokia;
  • Samsung/LG’s failed OS attempts;
  • the PalmOS saga and many others.

Still, I believe Facebook has a chance here, and I will explain how.

Meet FacePhone

First of all, it’s a no-brainer, but it goes without saying that a potential Facebook Phone should not reinvent the wheel, and it should be based on Android. There’s no other open-source operating system with an app ecosystem as vast as Android’s; this was the main reason that even Microsoft ceased its Windows Phone operations in 2017 and considered bundling its OS as a virtual machine inside Android.

Instead, Facebook should enhance the user experience to a point where it’s rich and in line with its core social mission. In other words, Facebook should invent a new relationship-centric user interface for mobile.

For the tech-savvy out there, one could do this in Linux by rewriting or modifying Gnome or KDE. I did this in my Linux distro, Turkix, fifteen years ago with SimpleKDE, which tried to replicate the Windows look and feel on the Linux kernel.

In Android, this would require modifying the stack, namely, Window Manager and View System.

But how?

FacePhone should be organized around people and groups, rather than apps. In linguistic terms, the phone’s first-class citizens should not be “verbs” but “objects.”

For the last decade, a typical mobile user interface has looked like this:

with many many verbs as “apps”.

With FacePhone, the apps would be replaced by your relationships; groups (aka family, work, hobbies & interests, fetishes, etc.) and friends, similar to the mockup below:

where the first icon gets you to your contacts list (full list of your contacts in a list format) and the second one to the asocial apps — in increasingly fewer numbers — such as weather and calculator. The asocial apps category would also provide backward-compatibility for Android apps that have not transitioned to Facebook’s new social app direction.

When you click a relationship icon, it would list all the social interactions you can take, like send a Snapchat story, a DM on Messenger, or a call on WhatsApp.

It would be possible to enhance the relationship selection process with the multi-gesture features on mobile devices. For example, one can use multi-touch to select two groups and act on them (such as send a message).

Now to illustrate these concepts, I will show you some more realistic pictures with a phone with the user’s main social circles such as his family, alumni network, work and hobby groups with the Stanford network selected to call, text, voice message, list members, etc.

This is similar to a concept we created back in 2012, as part of our Altay-OS efforts to augment the Grou.ps vision in mobile. Please see the mockups and videos I created back in the day. It gives you a good (and amateur-ish) overview of what I have described above.

Is it patented?

While researching for prior art, I was able to find a few similar concepts patented in the past, one from T-Mobile. They are:

  1. Relationship Centric Mobile Interface: https://patents.google.com/patent/US20130145280A1/en?oq=20130145280
  2. Preferred contact group centric interface: https://patents.google.com/patent/US20130145280A1/en?oq=20130145280

The second patent is almost expired.

Previous similar attempts

Facebook did try a Facebook Phone in the past with failed results. For example, in partnership with HTC, it launched HTC First with a lot of fanfare, but like most other hardware endeavors of Facebook, HTC First hardly sold. The phone would come with a fork of Android that would be “deeply social,” but in my opinion, it lacked the imagination to be considered social.

Also, there was a start-up called Humin, which came up with the interesting concept of organizing your contacts list with AI. Humin was rumored to be acquired by Facebook, but it never came to fruition. However, Humin’s technology (or others like it) definitely would be interesting to generate the list of circles to show on one’s home screen.

How to program?

To boost developer interest in Facebook’s new Android-based device, Facebook would have to separately create a new low-code/no-code environment. Having already built successful open-source developer products like React, React.Native, HHVM, and GraphQL, Facebook is most certainly no stranger to this. Our very own, MIT-licensed Phở Networks represents a good prototype of what a social-first programming environment could look like.

As you can see from the image above, it comes with a low-code tool that lets you define your apps’ server-side components in a whiteboard-friendly fashion, using nodes and arrows (edges) to define your app graph-natively. For more information, check out this tutorial below:

Please note, one could still program on FacePhone using Kotlin or Android Java, but these apps would go to the asocial category — that’s unless the social specs are declared in a manifest file.

Is it the right approach?

Yes, it is. It is a head-on attack against Apple, and it should be. Apple is attacking Facebook nonstop with its privacy motto, new family of social apps, and even Memoji. It controls the most lucrative gates (the iOS) through which Facebook reaches its consumers. As Mark Zuckerberg stated, Apple is now their biggest competitor. While Oculus is a nice endeavor, it has a decade to become the mainstream way to consume games and movies. Furthermore, it doesn’t even signal that mobile phones will go away anytime soon. Therefore, Facebook should act or risk going extinct like another digital media property of the past, Yahoo.

With $70 billion a year in revenue and an operating margin of 35%, Facebook has all the resources and incentives to experiment in strategies to thwart the Apple threat. Hence, they should greenlight all FacePhone projects on their desk.

BONUS: an open-source FacePhone?

Do we really need Facebook to launch a FacePhone or can we build one in open-source?

Well, because Facebook hosts the largest virtual social graph in the world, yes it is the ideal organization to host an endeavor like this.

But one can start this by either forking KDE’s Plasma Mobile (which is still in its infancy) or editing Android’s source codes.

Either way, if you’re interested in working a project like this, please let me know. My email address is sokullu [at] protonmail [dot] com

The Startup

Get smarter at building your thing. Join The Startup’s +737K followers.