How to Measure Company Culture: Methods That Work

'Culture cannot be measured' is one of the most persistent misconceptions in HR. Culture itself is indeed not directly measurable, no more than your health is directly measurable. What you measure are indicators: signals that move with the underlying state.

The real question is not whether you can measure culture, but which frequency and level of detail produce something you can act on.

The problem with the annual survey

The default approach is one large engagement survey a year. Forty to sixty questions, a participation rate that slips a little each year, and a report that appears months later.

Three things go wrong:

The lag. A problem that starts in March lands in the November report and gets discussed in January. By then the people affected have either left or adapted.

The average. A single organisation-wide number hides exactly what you need to know. An overall 7.4 could mean everyone is reasonably content, or that four teams are excellent and one is in free fall.

The length. By question thirty, people are answering to finish, not to say something. Long questionnaires buy detail with quality.

Methods that do work

eNPS — one question, quarterly

The Employee Net Promoter Score asks how likely someone is to recommend the workplace, on a scale of 0 to 10. Promoters (9–10) minus detractors (0–6) gives a score between −100 and +100.

The strength is not the number but the simplicity: one question can be asked often without people dropping out, and that makes the trend line usable. Measured once a year it is a figure. Measured quarterly per team it is an instrument.

The open follow-up — why that score — almost always yields more than the score itself.

Pulse questions — one topic at a time

One short question about one specific aspect: collaboration this week, clarity of priorities, workload. Rotate the topic and in a quarter you build a picture an annual questionnaire never reaches, at a fraction of the effort each time.

Day ratings — the trend, not the moment

One click on an emoji says nothing about a single day. Across eight weeks and per team it says a lot, because you see direction. A team declining three weeks running is worth a conversation regardless of the absolute level.

This is precisely the signal an office gave you free and that disappears when you go remote.

Behavioural indicators — what you already have

The most honest culture measurement often sits in data you already hold: how long new hires stay, turnover per team, how often recognition is given between colleagues, how quickly problems get raised. Nobody fills in a form for those, and nobody can answer them in a socially desirable way.

The trap that makes most measurement worthless

It is not the method. It is the silence afterwards.

If people answer and visibly nothing happens, they learn within one round that the survey is theatre. Next time participation drops, and the people still answering are not the people whose answer you needed.

So the one indispensable step is the response: what we heard, what we are doing, and what we are deliberately not doing and why. The third matters most. An organisation that explains why it is not acting on something gets believed. An organisation that promises everything and delivers nothing gets believed once.

A practical starting point

If you were starting today, in this order:

• One pulse question per week per team. Short, rotating topic, always with visible follow-up.
• eNPS quarterly, broken down per team. Never look at the organisation-wide average alone.
• Explain anonymity honestly. Say what you can and cannot see. In small teams even an aggregate can narrow things down, and people know it — pretending otherwise costs you exactly the trust you needed.
• One visible action per round. Actually changing one thing does more for the next response rate than ten intentions.

Measurement does not change culture. But without it, culture change becomes a contest of opinions, and that is always won by whoever talks loudest.