I've been the CTO of Trendoline for almost a year. The company is at a crossroads and I have enough distance now to look back at the year honestly.
What went well
The technical execution was solid. We shipped a working product on iOS and Android, kept the infrastructure stable, iterated based on feedback, and kept the codebase in a state where changes were possible without everything breaking. For a one-engineer shop, that's not nothing.
I also got better at communicating technical decisions to non-technical stakeholders. That took longer than it should have but it improved.
What didn't
I waited too long to raise hard questions about whether the product was finding its users. I kept assuming the distribution problem would solve itself once we added the next feature or improved the next flow. It didn't.
A CTO's job includes being honest about whether the technical work is connected to actual progress. I was better at building than I was at that question.
The decision I'd reverse
Early on I built the backend in a way that made sense for the features we had, but didn't account for how the data model would need to change as we learned more about what users actually did with the product. Migrating that later cost weeks.
The principle I took from it: build for change, not for your current assumptions. The current assumptions are always wrong about something.
What comes next
I'm leaving Trendoline at the end of next month. The company and I are going in different directions and that's the honest framing. I'm moving to San Francisco to be closer to the entrepreneurial ecosystem I want to be part of. The Draper network is based there, and being around that kind of density of early-stage thinking is the right next environment.
I'm going to San Francisco.
With gusto, Fatih.