At some point, every startup hears the same sentence: “How do I get a raise?” And for the first time, there’s no really good answer.
That’s usually the moment people start thinking about HR structures. About a salary band. A career model. A clean recruiting process. The instinct is right. It just tends to point at the wrong template.
Because the picture in your head usually comes from the last big company you worked at. Twelve career levels, a competency matrix, calibration rounds. And you’re thirty people. A system built for five hundred suffocates a company of thirty long before it fits.
Five questions, one pattern
These systems almost all arrive through the same kind of moment. Through a question that no longer answers itself.
- Pay: “How do I get a raise? And when?”
- Career: “Where do I go from here?”
- Recruiting: “What are we actually looking for in new people?”
- Performance: “Are you happy with my work?”
- Culture: “What do we say no to?”
While the company was small, every one of these answered itself. Pay was a negotiation. The way up was obvious, because the company grew faster than people’s ambitions. The network sorted the candidates. Performance was a shared feeling. And culture was lived, not explained.
Each of these informal mechanisms worked until it stopped reaching everyone. That’s the real trigger. Not size in itself, but the moment the quiet machinery stops carrying the new people. From the outside you see it quickly. From the inside, in the middle of growth, usually last.
What a system is at this size – and what it isn’t
At thirty people, a system isn’t bureaucracy. It’s the standard everyone has been sensing all along. Said out loud for once. Or written down.
The difference lies in the purpose. At two thousand people, a system has to produce fairness where nobody can see the whole picture anymore. At thirty, you can see it. So the system doesn’t have to make people comparable. It just has to answer one question well.
In recruiting, that doesn’t mean an interview guide with scripted questions, a scoring grid, behavioural anchors per level. It can mean five points you agree on, which every hiring manager brings back from their conversation. Then you sit down together, compare what you observed, and decide.
Same with pay. It doesn’t have to be bands built on market data from three sources. It can be a logic that you can explain in a few sentences — even if it departs from the market. Some companies deliberately keep the differences between roles small, because belonging matters more to them than the last percentage point of market rate.
The room to move is bigger in this phase than it will ever be again. What you write down now is the V1. It might hold for a year, then you look at it again. And update it with what you’ve learned in the meantime. That takes the pressure off getting it perfect forever on the first try.
System: same word, different tool.
Three observations from practice
- A system is an answer, not an insurance policy.
The good ones come into being when a question slowly reaches the table — not years before. Built too early, they choke on a complexity nobody needs. Built too late, you’re putting out a fire instead of answering a question. “Ahead of the curve” means hitting the window in between. Hearing the question before it burns. - The test isn’t “is it complete?” but “can I explain it?”
If a pay logic can’t be explained in a few sentences, it’s too complicated. Explainability isn’t the nice extra. It’s the whole point. A large company can carry a system that in the end only HR fully understands. At thirty people, explainability is precisely the advantage. - A system doesn’t replace judgment, it sharpens it.
In large companies, systems also exist to depend less on individual judgement. At this size it’s the other way round: built to judge better, not to hand the judgement over.
So, now what?
The crises of growth are predictable. We’ve written about that before. So are the systems that answer them. The mistake usually isn’t building too late. Sometimes it’s building too big.
“Ahead of the curve“, then, doesn’t mean forcing a system onto the company that it might want in five years. It means seeing which question is about to lose its answer next. And being ready to build the right one exactly then. Not the biggest one. The one that fits, at the right moment.
And if the answer isn’t clear yet: that’s usually the point where we at TrustWork come in. As a pragmatic observer from outside, one who has seen the curve a few times. No big programme, no battle of concepts. Just the next question, looked at together, with some structure.
