When it comes to platforms, it has been well understood – the best platforms start out as applications.
A platform without a “killer” app, is like a tree falling in the forrest. No one knows and no one cares.
As a result platforms are born with “app first"
Facebook’s platform, F8, was born after it had millions of engaged users.
Apple’s iOS platform came after the iPhone success.
Developers built on top of Twitter after Twitter had traction as an application
So a fairly typical reaction when startups pitch a platform idea to prospective employees and investors is: "what’s the killer app to get this thing going?”
Yet we are also seeing valuable platforms emerge without the platform creator developing the very first killer app.
Few examples:
Yodlee for banking apps
Twilio for voice and sms apps
Cardspring for payment card apps
Clever for education apps
Onswipe for touch web apps
These “platform first” companies are getting traction because
a) they are solving an extremely messy (understatement) problem behind the scenes
b) their APIs are simple & clean. the ease of developing apps on these platforms are capital efficient (e.g. compare this with the old days of mobile app development pre iOS)
c) they are supporting and creating an ecosystem to flourish where the previous ecosystem was broken. The apps want the platform to succeed because the pain and mess previously felt was unbearable.
d) roles are fairly well understood between platform and developer. everyone leaves a little money on the table on both sides to get to a happy place.
I still like the idea of app first, platform second. In the best case, the app provider that has platform ambitions can dogfood the same API that they give developers.
But it’s no longer the only way to get a platform started.
It’s interesting to think about other markets where a “platform first” company could emerge