Small by design

Flagon is built out of small, autonomous teams with a lot of ownership and not much process in the way. This page explains how that's put together, and why we plan to keep it that way even as the company grows.

Ownership that's whole

The unit of work here is a small team that owns something end to end. Not a team that owns a slice of a thing and has to file requests with three other teams to change it, an actual team that can take a problem from "this is broken" to "this is shipped" without a relay race.

Small teams are faster, but that's not the main reason. The main reason is that ownership only feels real when it's whole. If you can point at something and say "we made that, and we can fix it," you care about it differently than if you owned a fifth of it and shared the blame for the rest.

Generalists who care

We hire generalists who care more than specialists who don't. The person who'll wander outside their lane to fix the thing that's bugging them, who reads the values and actually means them, is worth more to a small team than a narrow expert who stops at the edge of their job description.

That doesn't mean nobody has depth. It means the default posture is "I'll figure it out" rather than "not my area." On a small team, the gaps between roles are where things break, so we'd rather have people who instinctively cover them.

Minimal process, high ownership

Process is what you add when you've stopped trusting people, and most of it is scar tissue from a problem that happened once. We keep the minimum that actually helps and resist the rest. The test for any new process is simple: does it help the people doing the work, or does it just let someone at a distance feel in control?

Decisions go to the work

Decisions belong with the people closest to the problem. The person doing the work usually has the most context, so that's usually where the call should be made, not escalated up a chain to someone reading a summary of a summary.

That puts a real responsibility on everyone: if it's your call, make it, write down why, and own the outcome. We don't want decisions pooling at the top, and we don't want them dodged. For how this actually works in practice, see how decisions get made.

Leadership is a role, not a rank

Right now, leadership at Flagon is one person. Chase founded the company, and for the moment he's also the engineering, the product, the design, and the support, plus whatever else the day needs. That's not permanent, but it's the honest state, and it shapes what leadership even means here.

We're aiming for something closer to Valve than the usual pyramid. A leader here sets direction, holds the quality bar, and clears the path so other people can do their best work. What a leader doesn't get is rank that outranks the people doing the work, or the right to be escalated to instead of the person with the context. Nobody reports up to be told what to think.

As we grow, the goal isn't to add layers between people and their work. It's to keep the number of people any decision passes through as close to zero as we can, and to hand pieces of leadership to whoever is best placed to hold them. The flatness is the part we intend to keep.

Staying small on purpose

We intend to stay small even as we grow, which sounds like a contradiction until you look closer. Growing doesn't have to mean one big org with layers of management standing between people and their work. It can mean more small teams, each still owning something whole, each still able to move.

The failure mode we're avoiding is the one where a company gets big and slow at the same time, where coordination eats the day and nobody can point at a thing they clearly own. Staying small is how we keep the parts of being early that are actually good: speed, ownership, and the ability to change our minds without a committee.

We won't always get the balance right. When we're adding process that doesn't earn its keep, or letting a team grow past the point where ownership stays whole, that's a bug. Say so, on GitHub or at hey@flagon.io. Staying small is a choice we have to keep making, not one we make once.