Reflections on my first startup attempt

Reflections on my first startup attempt

“And so castles made of sand fall in the sea, eventually”

-Jimi Hendrix

I started a startup, and I’m no longer part of it. Needless to say, this was not the exit I imagined.

Coming out of a strong fundraise, we had money in the bank, great investors on board, customers ready to pay to jump in, and had just closed the last candidate on our hiring plan for the year. We were in the midst of building a great product, but we had different ideas about where we stood, our velocity, and how to move forward. Rather than working through it together, my co-founders decided they wanted to part ways with me.

From the beginning, my biggest risk when we started the company was obvious: I was joining two people with similar backgrounds, in finance, who knew each other well. As the only technical one out of three co-founders, I was the odd one out on multiple counts. It was clear that I was stepping into a well known founder dynamic, but I convinced myself that the market pull and the specific people involved would offset that risk. In the end, it all boiled down to mismatched expectations.

Out of the lessons of this past year, my biggest learnings were about myself:

I also learned a few lessons along the way that can hopefully help other aspiring founders, and not just my future self:

I am convinced that my early departure was for the better. I wish my former team all the best.

Eventually, I’ll put this experience to good use and take another swing at the startup piñata. In the meantime I am open to fractional engineering roles while I find my next full-time opportunity at a more established startup. If you’re looking for a software or data engineering leader to join you, or simply want to hear a founder’s perspective, hit me up.

Photo: Photo: “Peace Out,” Bangkok, Thailand, by me. Previously posted on Thailand, 2023.

  1. NetSuite replication is a notoriously hard problem. This is in part because of their clunky APIs, and in part because of how customization interacts with permissioning. This is why vendors like Merge and Rutter offer a limited set of objects, and drop out to a proxy instead of solving the full problem for their users. If you need a full ledger replica, and want to support custom objects, you’re forced to build your own. 

  2. This often misquoted fact comes from the opening pages of Noam Wasserman’s The Founder’s Dilemmas. He in turn cites 65% from a 1989 paper by Gorman and Sahlman which points to “senior management as the most important contributing factor” to company failure, but “senior management” is not necessarily the same as co-founder conflict. To make things worse, the paper is a survey of VCs in 1984! More recently, there’s anecdata from Elad Gil in 2013 quoting a YC founder who blamed failure of ~20% of his batch to co-founder issues. In 2012, Paul Graham wrote on HN that ~1/4 of companies lose a co-founder, but that’s also not failure. I find it odd that there aren’t better sources for this. 

Want to see more articles like this? Sign up below: