There's a pattern almost every startup follows. In the early days, "HR" means one thing: hiring. Once you have a team, you trust people to figure things out. Then somewhere between employee 8 and employee 15, something breaks — a performance issue nobody knows how to handle, a dispute about leave, a resignation from someone who felt undervalued and never said so. And suddenly the founder is spending three weeks on people problems that a basic process would have prevented.
The issue isn't malice or incompetence. It's timing. People ops infrastructure feels unnecessary until the moment it's urgently necessary — and by then, you're building it under pressure.
Here's a stage-by-stage framework for what to put in place and when.
Stage 1: 1–5 employees — the fundamentals
At this stage you probably don't need an HR system. But you do need three things in writing, because verbal agreements create disputes and unwritten norms become assumptions.
Stage 2: 6–15 employees — where most startups stall
This is the danger zone. The team is big enough that informal coordination no longer works, but most founders haven't made the mental leap from "we're a small team" to "we're an organisation that needs operating infrastructure."
The specific problems that surface in this range:
- Two people on the same team have different understandings of their performance expectations
- Someone is given a raise; others hear about it; nobody understands why
- A manager-level employee joins and has no context for how decisions get made
- Leave records are inconsistent across the team because tracking is ad-hoc
Stage 3: 16–25 employees — intentional culture
At this size, culture stops being something that happens organically and starts being something you actively manage — or fail to. The founders are no longer in every meeting. Managers are starting to make decisions on behalf of the company without direct oversight. New hires are being shaped by the team culture they encounter, not by the founders' intent.
The most common mistake: waiting for a crisis
The trigger for most founders to finally build people infrastructure is a crisis: an unexpected resignation, a team conflict that escalates, a compliance audit, or a failed hire that cost six months of salary. At that point, they build it — but under pressure, reactively, and often with trust already eroded.
The better approach is to treat people ops like product engineering: ship early, iterate, don't wait for perfect. A simple leave policy written today is infinitely better than a comprehensive one written the week after a dispute.
The specific things to prioritise, in order of impact:
- Written agreements before verbal commitments go stale. Memories diverge fast.
- A system that removes manual reconciliation. Spreadsheets are a coordination tax you pay daily.
- One feedback touchpoint per quarter per person. This alone prevents most surprise departures.
- A documented process for the things you dread. Performance improvement, offboarding, pay decisions — write the process before you need it.
People ops isn't about adding bureaucracy. It's about removing ambiguity. Every process you put in place answers a question your team is currently answering informally — inconsistently, invisibly, and often in ways that don't reflect your actual values.
Workived gives small teams a single place to manage employee records, leave, attendance, documents, and org chart — all in one tool, free for up to 25 employees.