January 7, 2026

I spent 2025 sitting across from the founders of Sentry, WorkOS, Resend, PlanetScale, Chroma and Mastra in San Francisco. Face-to-face. They’ve built unicorns, category-defining devtools, and they told me what actually works.

Irina Nazarova CEO at Evil MartiansBook a call

Go get a job

Pressure to succeed has never been higher. Founders are trending younger. YC’s median age has dropped from 30 to 24 in two years and applications from 18-22 year-olds are also up 110% year-over-year.

But David Cramer, founder of Sentry, has advice that goes in the other direction: “I actually think the biggest mistake people make is starting a company versus just joining one … Sentry is my sixth job and I started full-time. I think I was 30 years old. And I learned a lot from those experiences. And then you have like 20-year-old starting companies. You just don’t know anything yet. You haven’t experienced enough. You haven’t seen everybody else mess it up yet.”

Talk to 100 users

Engineers love to build. That’s the fun part. The hard part? Talking to users before you’re sure the idea is good. Instead, most founders exhaust every other option first: more features, better docs, one more refactor.

Michael Grinich, founder of WorkOS, says: “Engineers have the tendency to exhaust every other possibility before talking to customers. We like building stuff … Nothing destroys a good flow state like talking to a user and having them tell you like, I don’t know if this thing is really useful.”

Get to “No” faster

Building new features? “That’s nice” is not the feedback you want. You want: “I need this today, can I pay you for this?” Hearing a “no” is the uncomfortable truth that gets you closer to PMF.

Zeno Rocha, co-founder of Resend, says: “I want the market to tell me if this is a product or if this is a feature. So it’s almost like detaching yourself from your beliefs and just trying to seek the truth. I’ll let the market tell me if this is good or or bad, and I’ll try to cut scope as much as possible.”

Customers don’t want to think about you

The best SaaS becomes part of your workflow so seamlessly you forget it ever existed.

Sam Lambert, CEO of PlanetScale, says: “The magic phrase we hear from people is, I just don’t wanna think about my database anymore. Now that is a goal.”

Consensus kills great products

Design by committee produces mediocre products. Small, opinionated teams win.

Jeff Huber, co-founder of Chroma, says: “You can just reference Linux itself, right? You know, it is maintained by a small group of highly opinionated people. This is why it is so great. If it was one large decision by committee, it would be a trainwreck.”

Generosity breeds luck

Building community is the long game. That means giving before asking.

Abhi Aiyer, CTO and co-founder of Mastra, says: “We learned from Brad Flora at YC. That generosity breeds luck … We will do whatever. You want to meet somebody, you want to hang out, are you feeling bad? … And that’s been a game changing difference…”

Building for developers in 2026

Some lessons stay the same no matter the era: Talk to your users. Focus on DX. Ship fast. Stay opinionated. Let’s build some amazing tools for developers in 2026.

New episodes of Dev Propulsion Labs drop every two weeks. Subcribe here.

Irina Nazarova CEO at Evil Martians

Each year we work with 40 developer tool startups, like bolt.new, Wallarm, Lingo.dev, Steel and Synadia. Preparing to fundraise? Let’s talk about your strategy!