A Growing Company Decision Record: From 20 to 200 People

Scaling decision-making in a growing company means changing how decisions are recorded at each stage: verbal consensus works at 20 people, a lightweight log is needed by 50, and a durable, queryable decision record becomes essential by 200. The failure is gradual — hallway alignment stops scaling long before anyone notices the institutional memory has evaporated.

What breaks as you scale

At 20 people, everyone hears everything. Decisions happen in the open, context is shared by proximity, and nothing needs to be written down because the whole company fits in one conversation. This works so well that founders assume it always will.

It does not. Somewhere between 50 and 100 people, the company crosses a threshold where no individual holds the full context anymore. Decisions made in one team are invisible to another. New hires inherit no history. The same questions get re-argued because the people who settled them are now three layers away. By 200, the cost of not having a decision record is a daily tax — the coordination tax on distributed teams compounding across the whole org.

Decision practices by company size

Each stage of growth demands a different decision practice. The mistake is staying in an earlier stage's habits too long.

Stage	What works	What breaks
~20 people	Verbal consensus, shared context by proximity	Nothing yet — but no memory is forming
~50 people	Lightweight decision log, named owners	Cross-team decisions start going dark
~100 people	Structured records with rationale and authority	Re-argued decisions, onboarding amnesia
~200 people	Queryable system of record, representation layer	Without it, coordination tax dominates

The distinction between a casual log and a true record matters more than it sounds — it is the difference covered in decision log versus decision record.

The transition companies skip

The transition most companies skip is the move from documenting decisions to making them durable and queryable. A decision log that no one can search, that does not capture authority, and that cannot answer a question when the owner is offline is barely better than no log at all. The leap is to a record that anyone — including an AI assistant or a new hire — can query without finding the original decision-maker.

This is the point at which decision-making becomes infrastructure rather than habit. Formalizing it under a decision governance framework gives the growing company a consistent standard: what gets recorded, who has authority, how reversible a call is, and where the audit trail lives. Companies that skip this transition feel it as a mysterious slowdown — everything takes longer and no one can say why.

The scaling playbook

1. Start the record before you need it. The cheapest time to build decision infrastructure is at 30 people, not at 150 in a crisis. Begin while the volume is manageable.
2. Capture authority as you add layers. Every new management layer creates authority ambiguity. Record who decides what so the org chart and the decision map stay aligned.
3. Standardize the decision record. One format, one location, queryable by everyone. Fragmented wikis recreate the problem at larger scale.
4. Adopt silence over speculation early. Make it a cultural norm before headcount makes guessing dangerous: if it is not recorded, it is not decided.
5. Build the representation layer at 100+. When owners are unreachable and teams are global, the record must answer in their place to keep work moving.
6. Onboard via the record. New hires should learn the company's decisions by querying the record, not by interviewing tenured employees who become bottlenecks.

Who owns this as you grow

Below 50 people, decision infrastructure is everyone's casual responsibility and no one's job. As you scale, it needs an owner — usually the chief of staff or a COO. The chief of staff typically builds and maintains the system; see the chief of staff guide to decision infrastructure. The COO uses it to coordinate without drowning in meetings, covered in the COO playbook for coordination without more meetings.

The throughline across every stage: as the company grows, the share of decisions that can safely live only in someone's head drops toward zero. Building the record early is the cheapest insurance a scaling company can buy.

Common Questions

How should decision-making change as a company grows?

Decision-making should move from verbal consensus at around 20 people, to a lightweight log by 50, to a structured record with rationale and authority by 100, and to a queryable system of record by 200. Each stage requires a more durable practice because no individual holds the full context as headcount grows.

When should a growing company start recording decisions formally?

Start before the pain peaks — around 30 to 50 people, while decision volume is still manageable. Building decision infrastructure in a crisis at 150 people is far more expensive than establishing the habit early.

What is the difference between a decision log and a decision record?

A decision log is a casual list of what was decided; a decision record is durable, queryable, and captures authority, rationale, and reversibility, and can answer questions even when the owner is offline. The leap from log to record is the transition most scaling companies skip.

Who should own decision infrastructure in a scaling company?

Typically the chief of staff builds and maintains it, while the COO uses it to coordinate operations. Below 50 people it is everyone's loose responsibility; past that, it needs a named owner to avoid the institutional memory eroding as the org grows.