Every plan we have ever written had a headcount number in it, and every quarter that number has been wrong on the high side. Not because we could not afford the people. Because we kept deciding, deliberately, not to hire them yet. I want to explain that decision honestly, including the parts where it cost us, because most writing about hiring treats speed as an unqualified good and I think that is one of the more expensive ideas in company building.
The pressure to hire fast is relentless and it comes from everywhere. A team is underwater, so the obvious fix is more hands. A competitor raised a round and is posting forty roles, so it feels like falling behind. A board deck looks healthier with an aggressive ramp than a flat line. All of that pressure points the same direction, and almost none of it asks the question that actually matters, which is not "can we add this person" but "will this person make the team better than it is today."
Speed in hiring is the easiest metric to fake
Headcount is seductive because it is legible. You can put it on a slide, watch it climb, and feel like the company is growing. Recruiting velocity feels like progress in a way that almost nothing else does, which is exactly why it is dangerous. It is the easiest number to move and one of the worst to optimize for, because the cost of a bad hire does not show up on the chart that celebrated the hire.
The math here is not subtle once you actually run it. A mis-hire at a senior level does not cost you their salary. It costs you the salary, plus the months of management attention spent trying to make it work, plus the opportunity cost of the role sitting effectively empty while occupied, plus the morale tax on everyone who has to compensate. Across a meaningful role I would put the all-in cost of a wrong hire at somewhere between one and two times annual compensation, and that ignores the worst case, which is the bad hire who is good enough to keep and bad enough to slowly lower the bar for everyone after them. That last one is the most expensive thing in company building and it never appears in any model.
Against that, what does slowness actually cost? An open role. A team that runs hot for another month. A roadmap that moves a little slower. Those are real costs and I do not want to wave them away. But they are recoverable. A bad hire embedded in a small team is often not.
A small team of people who fit compounds
The reason we tolerate the open role is that a small, well-fit team does not add linearly, it compounds. When everyone trusts everyone else's judgment, you can skip enormous amounts of process. Decisions happen in a message instead of a meeting. People take real ownership because there is nowhere to hide and no one wants to hide. Context does not have to be rebuilt for every conversation because everyone already shares it.
Add the wrong person to that and you do not get a slightly larger version of the same thing. You get something qualitatively worse. The trust that let you move fast now has a gap in it, so you start adding the process that the trust used to make unnecessary. The reviews get heavier. The documents get longer. The meetings come back. People route around the weak link, which is its own kind of overhead, and the whole system gets slower precisely because you tried to make it faster. A ten-person team can out-build a thirty-person one for exactly this reason, and it is often not close.
Example: Imagine a team so far behind on a launch that it very nearly hires two engineers in a week to brute-force the problem. Suppose instead it pauses, ships late with the people it has, and uses the slip as the forcing function to fix how work gets scoped. A year and a half later that same small team, barely grown, is shipping more per quarter than the bloated version would have. The discipline forced by staying small turns out to be the actual asset. That pattern repeats often enough that I trust it.
What we look for, and why it takes longer
Hiring slowly is not the same as hiring lazily. It is more work, not less. We are not waiting around. We are running a higher bar, and a higher bar takes more candidates, more conversations, and more willingness to say no to someone who is good but not right.
The two things we weight most heavily are also the two that are slowest to assess. The first is whether the person raises the average. We use a simple, brutal test: would this person make our existing team want to work harder, or would they make our best people quietly start to disengage. You cannot answer that from a resume. It takes real interaction. The second is whether they fit the way we work, which for us means remote, asynchronous, and writing-heavy. Someone can be excellent in an office culture and miss completely in ours, and there is no amount of talent that compensates for not being able to operate the way the team operates.
Both of those take time to evaluate well, and both are easy to skip when you are in a hurry. That is the whole point. The pressure to hire fast is precisely the pressure to stop checking the things that matter most.
There is a third thing we have learned to weight, and it is the one most teams get wrong: we hire for the company we are, not the company we wish we were. It is tempting, especially when you are recruiting senior people, to bring in someone whose experience comes from a much larger and more structured place, on the theory that they will help you grow into that. Sometimes that works. More often the person spends six months frustrated that the scaffolding they are used to does not exist, the team spends six months frustrated by process they did not ask for, and everyone concludes the fit was wrong when really the timing was. We try to hire people who would be excited by the stage we are actually at, not the stage we will be at in two years, because the person has to do their best work now, in the conditions that exist now, not the ones on a slide.
The empty seat is a feature, not a failure
The mindset shift that made this sustainable was learning to treat the open role as a position we have not filled yet rather than a problem we have failed to solve. Those sound similar and they are completely different. The first keeps the bar intact. The second creates urgency that quietly lowers it, because once a role is a "problem," any plausible candidate starts to look like a solution.
We protect against this in a few concrete ways. We never let a hiring manager interview alone for a role they are desperate to fill, because desperation is the enemy of the bar and it is invisible to the person feeling it. We write down what "great" means for a role before we start, so that nobody can quietly redefine it downward at the end of a long search. And we make it culturally acceptable, even expected, for a leader to come to a planning meeting and say the role is still open and that is fine. The leaders who can sit with an open seat without flinching tend to build the strongest teams, because they are the ones who never settle.
I will be direct about the trade-off, because pretending it does not exist would undercut the whole argument. This approach loses things. There are projects that would ship sooner with more people. There are stretches where a lean team gets run harder than anyone would like. Hiring slowly is a real bet with a real cost, and anyone who tells you their approach has no downside is selling something.
How we keep a small team from drowning
The honest counterargument to all of this is that a small team can break under load, and a broken team is not a strength. That is true, and it is why hiring slowly only works if you are equally serious about removing work the team should not be doing in the first place. A lot of what feels like a hiring shortage is actually a tooling and process problem wearing a hiring shortage costume.
Before we open a role now, we ask a harder question first: is this genuinely more work than a reasonable team can do, or are we asking the team to spend its hours on things software should be doing. Consolidating our own stack onto Atlas so that tasks, projects, inbox, and automations live in one place removed a surprising amount of the busywork that used to masquerade as a need for more people. The cheapest hire is the one you did not need to make because you stopped making the team fight their own tools.
So we keep hiring slowly, on purpose, even when it costs us, because the alternative costs more in the currency that actually matters. A company is not the number of people in it. It is what those people, together, are capable of, and the fastest way to lower that ceiling is to fill the room in a hurry. We would rather have the open seat than the wrong person in it, and almost every time we have honored that, the team has been glad we did.