Great output isn't big moves. It's the lunches.

A friend of mine runs a small, YC-backed startup in London, and I got to spend a full day at their office. Just heads down, hanging out, watching how a genuinely high-output team works up close. I went in expecting to learn something about product or strategy. What stuck with me was smaller, and honestly more useful.

It was all in the details. And underneath every detail was the same thing: intent.

My friend is the founder, and you can feel it in every decision he makes. Even the tiny ones have an afterthought behind them. Nothing runs on autopilot. That sounds obvious, but watch a team for a day and you notice how rare it actually is. Most decisions in most companies happen by default. His happen on purpose.

Two founders, total alignment

Here's the one that made it click. The two founders are about as aligned as two people can be. They're close, in constant contact, doing basically everything together. That sounds intense, and it is, but here's why it matters: founder and co-founder drifting apart is the single biggest thing that kills a startup. Not the market, not the product, the two people at the top falling out of sync. These two have engineered that risk out. They can settle anything over breakfast, because they're never out of each other's orbit to begin with.

Is there a version where that closeness is too much and it burns you out? Of course. But they've clearly made it work, which says as much about the relationship as it does about the discipline.

How he hires: barely at all

The same intent shows up in how he hires, which is barely at all. They're well-backed, and they haven't sprayed the money on headcount. He treats every hire as a serious decision and is ruthless about fit. He'd rather spend real effort and real money up front making sure someone is right than add the wrong person. Because he understands the actual math: a wrong hire isn't just a salary. It's his time, his focus, and the team's momentum, and that costs far more than being thorough ever could. Money in the bank is optionality. The wrong hire is a tax you keep paying.

Claude as a teammate, not a gadget

The tooling carried the same fingerprint. They've woven Claude straight into how they work, connectors into their email and their docs, and they talk to it the way you'd message a teammate in Slack. It's not a toy they pull out for a demo. It's load-bearing. It's not perfect, but it makes the work genuinely faster, and watching a small team lean on it that fluidly was its own quiet lesson. Funny timing, I'd wired the same kind of connectors into my own setup the same week.

And then, the lunches

This is the one that stuck. They keep the fridge stocked with proper, high-quality lunch. Good enough that people actually want to eat in. And every day the whole team eats together, same time, up on the balcony. That's it. That's the whole thing.

But sit with what it quietly does. Nobody scatters to grab food alone. The team eats together by default, so the camaraderie isn't an offsite you book once a quarter. It's just Tuesday. It costs a little money and almost no effort, and it builds more culture than most culture initiatives ever will. Nobody decided to "improve team bonding." Someone just decided the lunch should be good and shared, and the bonding came for free.

That's the through-line. None of it is accidental. The crisp roles, the lean setup, the careful hiring, the tooling, the lunches. There's intent behind every small thing, and you feel it the moment you walk in.

What I'm taking home

We love to believe great output comes from the big moves. The strategy, the funding, the headcount. From a day inside a team that's genuinely good, it looked like the opposite. Great output is a handful of small things, chosen on purpose and done consistently. The founder isn't thinking big thoughts all day. He's being deliberate about a few small ones, over and over, and the big stuff mostly takes care of itself.

I left thinking about all the small, intentional things I'd been treating as afterthoughts. The roles, the rituals, the defaults, the lunches. That's the part I'm bringing back to Marnix.