TL;DR

Culture exists to help talented humans do their best work. At N3TWORK it was our unfair advantage—until we treated Version 1 like a finished product. A handful of mis‑hires, loose progression pathways and a clear but not always respected hiring filter eroded the edge. Here’s what shifted and what I’d handle differently next time.

Where we started

In the early days the whole team could recite our five values—Greatness, Delight, Ownership, Dialogue, Action. Context flowed fast, decisions were made in Slack threads, and we shipped every week. Momentum felt inevitable because the cultural operating‑system, as we called it, matched the people using it.

How the cracks appeared

We always knew the values would need to evolve, but I underestimated how deliberate we’d have to be about revisiting and re‑aligning them as the headcount grew. As the team grew, I over‑indexed on raw talent and brought in a few individuals who, despite their brilliance, never quite worked the way we did. Because the progression framework was a single ladder, specialists and would‑be managers found themselves queuing for the same rungs, and the clean feedback loop we relied on began to fuzz.

Onboarding itself wasn’t the culprit; the miss was upstream. We didn’t screen rigorously enough for cultural fit, and when mis‑aligned players did slip through we hesitated to act. In hindsight we should have taken a page from Riot Games—move people on early, even pay them to walk if the fit isn’t mutual. Context that used to travel by osmosis slowed to a trickle. A quiet divide formed between those who understood the original playbook and those drifting on the edges. Leaders reacted by adding extra checkpoints, which dulled our pace. By the end of the quarter we managed about twenty per cent of the OKRs we’d set. Two promising game prototypes were quietly shelved, not because the ideas were poor but because decision‑making had lost its snap.

(It’s worth saying: even in this state our culture was better than most places I’ve worked. It just stopped being the accelerant it once was.)

What I’d do now

(Culture Version Control: deliberate, scheduled updates—because an OS only stays secure if you patch it.) First, act with intentionality—treat culture like an operating system under version control and push regular updates. I’d schedule a values review every six months. The principles wouldn’t change, but the reference stories and examples would stay current. I’d interview candidates for drive before skill—hunger ages well, craftsmanship can be taught. Onboarding would switch from static slides to immersive stories: shadow key meetings, founder AMAs, and a written reflection in week two so newcomers prove to themselves they’ve absorbed the lore.

Career progression would split into two clear tracks—technical depth or leadership breadth—each with equal status. Finally, I’d block a thirty‑minute weekly pulse check: a fast, honest look at what feels off and a decision there and then about who will fix it. Think of it as a quick weed‑pull before anything takes over the flowerbeds.

If you want to borrow anything

Write a one‑pager of non‑negotiables today. Identify any "brilliant but disruptive" people and act decisively. Map each teammate’s motivation—intrinsic or extrinsic—and coach accordingly. Draft those dual career tracks and circulate for comment. And, yes, put that weekly pulse check in the diary for Friday.