Biz Stone, the co-founder of Twitter, Medium, and Ask Jelly, explains what it takes to identify--and sustain--a business that reaches millions of users.

Inc.: How do you know when it's the right time to start a company or make a major career move?

Biz: I've always followed a person and never followed money or an insti­tution. When I left Google, it was because Evan Williams left. We did Odeo, a podcasting startup. Next, I followed Jack Dorsey, who's now Twitter's CEO. I was always linking myself to a person smarter than me. That's how I got ahead. There wasn't a master plan.

Is there a moment when you can tell your company is going to be a success?

For Twitter, the tipping point was in March 2007 at South by Southwest. There was an influential tech guy who had a bunch of followers on the prototype of Twitter. He was at a pub and it had gotten really loud, so he sent out a tweet that said, "Hey, anyone who wants to talk about projects, let's meet at this other place where it's quieter." In the eight minutes it took him to walk to that pub, it had filled to capacity, with a line out the door. It blew my mind, because I realized that we had invented something new. Two days later, literally, we formed Twitter Incorporated.

What was your biggest early challenge at Twitter, and how did you deal with it?

We made a lot of mistakes in the early days. We shot ourselves in every major organ and still managed to succeed. For one thing, we built the entire platform on a coding language that's not meant to handle lots of people. The site was breaking every day. Every time it went down, I'd explain on our blog why it happened, how we fixed it, and that it would never break in exactly that same way again. It almost became advertising for us. There's value in being vulnerable and radically transparent.

As an investor, what's your top piece of advice for startups?

What I usually advise people is to build a prototype just to see if your idea is a thing or not. When I invested in Slack, it was a video game company. Then they realized that all the internal communication tools they had built for their spread-out company could actually be the company. You have to start somewhere and see where it takes you.