Google's former CEO Eric Schmidt on the data strategy behind the next hundred billion dollar companies:
He explains that he evaluates startup ideas by looking at 5-year growth trajectories and asking whether there's a more scalable strategy available.
Take a founder building an app they want to charge $10 for.
@ericschmidt's challenge:
"Why not give it away for free and upsell users instead?"
But the real insight is his framework for predicting which companies will dominate — and it all comes down to data.
"5 years ago, I said publicly that the future will be apps that are on smartphones that use Google Maps, GPS, and do something useful. Now, what I should have said was Uber."
So what does he think will define the next wave of massive companies?
Systems built on Android and iOS, fast networks, and powerful machine learning, with a crucial data advantage:
"They're going to use the crowd to learn something."
He illustrates with this example:
Pay dermatologists $1 each to categorize skin samples. Feed that into a machine learning system. Then sell the diagnostic service back to them, because a system trained on thousands of experts will outperform any individual.
Schmidt summarizes the winning data strategy:
"You crowdsource information in, you learn it, and then you sell it. [This] is in my view a highly likely candidate for the next hundred billion dollar corporations."
The blueprint: Aggregate expert data at scale → train ML systems on that data → sell superior insights back to the market.
This is how the next generation of dominant companies will be built.