Notes:

The first company went bankrupt 7 years after I left. With hindsight and with the experience of the second company, I came to see that waiting even nine years was far too long. The market expectations were off by several orders of magnitude, and the window of opportunity had passed in the first case.

The actual revenue on the previous slide was $2M for the first company. It turns out that many of the first company's customers were not paying, because they were dissatisfied with the service. Instead of adapting to this feedback, we continued on the course we had set for ourselves.

The second company had "paid charter customers" at launch. We had no need for a beta. The customers liked the initial product enough to pay for it. We continued to revise the product every week as we found bugs and customers requested new features. The first company had yearly updates, if that.

Naturally, I was a more experienced CTO. That experience has led me to believe that I don't know how to write software until it is written. That may sound strange, but it was very true. We had a long list of requirements for both companies' products, but the actual implementation varied greatly from the spec. Many features never got implemented in the requirements for the second company, and other requirements were added as customers demanded them.

Some other lessons are that you have to either succeed or fail fast. Don't hang around losing money for years on the hope that something is going to change. You have to either pay for your way, or move on. I have worked for two well-funded startups, and the first company was one of them. In both of the well-funded startups, we lost money for too long before we realized we had no product. If you can't generate revenues, don't be in business.