Why The F@*K Didn't We Just Make An App?!?!

I dont really know. Maybe because

NOT ALL INNOVATION WILL HAPPEN BECAUSE OF A FUCKING TOUCH SCREEN.

Sorry. Got a little excited on that one. But seriously not every problem can be solved with a mobile app. Yes the market for apps is huge. But there are also an aweful lot of developers, and a pretty riduculos number of apps. Its kinda like the biggest lottery ever created. Get out your napkin...

My numbers and assumptions

• 1bn devices
• 3mm apps
• 100 apps per device
• 90% of the apps are free
• 0.1% of apps are winners
• 15bn in total mobile revenue next year
• 15 apps are actually used weekly
• 100K mobile developers
• 5 yrs of development between smartphones and 1bn devices
• 6.3bn in VC money for app development
• Ignoring language
• Ignore appstore 30% cut
• mostly estimates - some sources: [1], [2], [3], [4]

Some maths

• 1bn devices x 100 apps / 90% x $0.99 = $9.9bn total spent on installs

• $15bn total revenue / 1bn devices = $15 per device (ie user)

• $15bn total revenue - $9.9bn spent on installs = $5.1bn remaining for in app purchases

• $15 per user / 15 used apps = $1 spent per used app per user

• $5.1bn spent in app / 1bn devices x 15 used apps = $0.34 spent per used app per user in app

• $9.9bn total install revenue / 3mm apps = $3,300 per app on install

• $9.9bn total install revenue / 3mm apps x 0.1% winners = $3,300,000 per winner app on install

• $5.1bn total in app revenue / 3mm apps = $1,700 per app in app

• $5.1bn total in app revenue / 3mm apps x 0.1% winners = $1,700,000 per winner app in app

• 3mm app / 100,000 developers = 30 apps per developer

• 30 apps per developer / 5 years x 52 weeks = 7 weeks spent per app

• 1,000 apps x 0.1% chance of app winning x 7 weeks per app = 134 man years of development

• $3.3mm + $1.7mm per winning app / 134 years of development = $37,313 yearly salary

• $37,313 app developer salary / $20,480 minimum wage in California = 82% better

• $37,313 app developer salary / $98,000 average software developer in California = 38%

vs. hardware

• $500 per user spent on hw vs. $15 per user spent on apps (33x more)
• $500 spent per device vs. $1 spent per app (500x more)
• ~500 weeks per device vs. 7 weeks per app (71x more)

Software is commodity

I think at this point I can safely argue that the apps, the software, has become a commodity. A single install is worth next to nothing, there are thousands of people building thousands of apps that are exactly the same. And its pretty hard to find any one app in amongst the millions.

Hardware on the other hand is rare, it takes longer to build, its where a lot of the real innovation and enabling happens. But it suffers from one very notable weakness.

It takes longer to build hardware

An app makes 33x less money but takes 71x less effort to build. What happens when it only takes 50x less effort? or 25x? or parity, what happens when it takes only linearly more time to build hardware? or when you make more than linear the amount of money on a sale? What percentage of apps fail? I assumed 99.9% but by comparison what percentage of gadgets fail? I'd wager a lot less than that.

Android is a free and open source operating system that has been shown to run on all kinds of hardware. There are enough designers of hardware that the chip manufacturers will sell anything to anyone these days. The design tools for hardware are becoming commodity. And there are dozens of startups attempting to lower the bar to designing and building hardware [upverter] - it looks a lot like what happened to software in the 90s and early 00's.

We are on the cusp of something transformative

You can see the start of it with companies like Square, Jawbone, and Nike. Smaller, connected, sensor based hardware. Designed in months by small teams working closely with software developers. Its consumer targeted and just as marketable as a mobile app (and often tightly coupled to one).

We arent that far away from the one-man-team hardware app developers.

Or manufacturing thats ubiqutous enough that you can just click "print".

Open source "libraries" that save many orders of magnitude in development times.

And design tools that encourage reuse and collabortation above isolation.

Imagine a world where Android runs on everything you own, it was designed by some startup or basement developer, it was manufactured just-in-time and it was shipped to you by Amazon.

Why we didn't just make an app

Because we belive in parity. We think there will be a day when its just as easy to build something with atoms and electrons as it is to build something with bits. We saw an ecosystem developing around the idea of open sourcing not just software but hardware too. And we wanted to help push this into happening. We wanted to build a product that multiplies innovation, more so than is an innovation itself.

People can already print in plastic [makerbot], and bash bits in a weekend [project72]. How far away can electrons really be?

About me

Upverter is my 3rd startup. I dropped out of highschool, and then university, both times to run startups. I studied Computer and Electrical Engineering and UWaterloo, and then cut my teeth designing the worlds most powerful network inspection hardware. I left Sandvine to make designing hardware a little less hard, and I think that means collaboration and accessibility.