Prototype before you write the check

The familiar venture sequence is company, pitch, investment. It works when the company already exists and the risk is mostly about execution and market.

Some of the ideas we care about do not start that way. They start as a thesis: a gap in infrastructure, a product that should exist, a market that is being served badly. There is no pitch deck because there is not yet a company.

In those cases we research, then we prototype. We put the idea in front of users and design partners. We look for revenue, or the honest absence of it. Only then do we form a company.

This is slower than writing a check. It is also how conviction becomes something you can stand behind for a decade.