What I've seen across hundreds of ventures 

Five years and 50+ co-founding teams assembled later, the pattern I see most often in studio ventures that find traction is not that the founders have better values, just sharper standards applied earlier.

Three observations.

1. Culture is what you tolerate

Founders set culture by accident more often than they set it deliberately. The first time someone misses a deadline and there's no conversation, that's the bar. The first time a co-founder ducks a hard customer call, that's the bar. The first time someone says "I'll get to it next week" three weeks running and nothing happens, that's the bar.

Dima, founder and CEO of Warren AI, is one of the best examples I've worked with of setting the bar through behaviour rather than rules. In Warren's studio phase, he was working day and night. When a bug needed fixing over the weekend and the team member on point didn't respond, he'd fix it himself at 2am on Saturday. He never had to write a working hours doc, never had to give a speech about pace. The early hires either matched the standard or self-selected out within weeks. The bar was legible because he embodied it.

Compare that with the founders I see set culture by avoidance. They feel the gap, give one round of feedback in the first 1:1, then say nothing for three months, hoping things will course-correct. They almost never do. Three months of silent disappointment hardens into something neither side can come back from. A standard set early, even a hard one, is a kindness.

The strongest founders I work with are explicit about the bar in week one through what they do, not just what they say. Every other piece of culture flows from that.

2. Standards live in week one

The expectations you set with your first five hires harden into precedent. If your first hire ships in seven days, the bar is seven days. If they ship in three weeks and no one says anything, the bar is three weeks. Hire number twelve will operate against whichever number you let stand.

This is why founders who scale well over-index on the first month. Cursor's two-day onsite. Fyxer AI's weekly "what is the one thing you'll put unreasonable effort into to move our most important goal?" question. ElevenLabs putting new joiners on live customer calls within days. The consistent move is to compress time-to-real-work and use the early signals to set or reset the bar.

I've also seen, up close, what it costs not to do this. One of our portfolio founders, building radiation-resistant materials for space out of our deeptech portfolio, grew from 6 to 20 people in four months. No structured onboarding, no explicit standards on focus or meeting hygiene, just the assumption that the bar set with the first five would naturally transmit. It didn't. Three months later, the founder was removing the coffee machine from the office and putting "No Phones" signs on the wall in meeting rooms, because the culture that had quietly emerged was 30-minute coffee breaks every hour and phones out in every meeting.

That's not a culture problem. That's a standards-in-week-one problem, surfacing three months late. Once a tolerance becomes a precedent, the cost of resetting it is roughly an order of magnitude higher than the cost of setting it the first time.

3. OKRs at the seed stage are usually the wrong answer

I see a lot of pre-seed and seed teams cargo-cult OKRs from companies eight years and 800 people ahead of them. The result is typically a quarterly planning ritual that consumes a week of leadership time, produces a document no one revisits, and sets metrics that don't actually reflect what's working.

We've watched this play out enough times across our own operations at Founders Factory. We've now moved away from OKRs in favour of a single North Star Metric, with one or two priorities a week cascading underneath it. We now advise our studio ventures, that especially pre-PMF, you need one number that captures whether the business is working, and one or two priorities a week that move it. Everything else tends to make the bar fuzzier rather than sharper.

PostHog wrote about this honestly in their company handbook. They moved away from OKRs because their engineers were "agonising over finding the right metrics" and the metrics didn't track the actual work. They switched to something closer to a single shared priority per quarter, using the W Framework originally written up by Lenny Rachitsky, where leadership sets direction, teams set the plan, and the unit of accountability is "we'll ship X" rather than "we'll hit metric Y".

The deeper problem with OKRs at this stage isn't the framework itself. It's what it incentivises. OKRs reward the team for hitting numbers, which means the bar quietly becomes "make the metric look good" rather than "ship the thing that matters". At seed, you can't afford that drift. The bar needs to be the work.

The bottom line

You don't get to choose whether you set a culture in the first five hires. You only get to choose whether you do it deliberately.

Standards aren't a thing you bolt on after Series A. They're not an HR function. They're the most leveraged decision a founder makes, because they compound. The bar you set with hire number one becomes the floor for hire number fifty.

Set it early. Write it down. Hold it.

Next: Part 3 — Radical Transparency as an Operating System