How I work
Culture change is rarely hard because people don’t care. It’s hard because organisations guess.
They launch initiatives, run workshops, publish values, and hope behaviour follows.
My approach is different. I use small behavioural experiments to replace guesswork with evidence, then I scale what works into everyday habits.
This is the Culture Experiment Loop.
Culture and behaviour change, without the uncertainty
I’m not claiming to make culture predictable. Humans are humans, after all.
What I remove is the uncertainty of vague plans and generic activity.
Instead of asking, “Will this land?” I help your team to design small tests that answer, “Did this change behaviour in real work?”
That shifts culture change from belief to learning.
What I mean by "culture experiments"
A culture experiment is a small, low-risk change designed to test one behavioural shift in a real team, in a real workflow, over a short period of time.
It's meant to be practical. Not perfect.
A culture experiment is...
- Specific: one behaviour, one context
- Small: low admin, easy to try
- Fast: typically 2 to 4 weeks
- Observable: you can see if behaviour changes
- Useful: it produces learning even if it “fails”
What it isn't
- A pilot that runs forever
- A workshop disguised as an action
- A comms campaign with a shiny new slogan or a poster campaign in disguise
- A grand programme launched top-down
Download the Culture Experiment Loop Toolkit - a lightweight guide with canvases and prompts for running culture experiments in real work.
.png?width=2000&height=1489&name=CEL_reduced%20(1).png "CEL_reduced (1)")
Step 1: Spot the friction
We start with where work is harder than it needs to be.
Friction shows up as patterns like:
- Decisions bouncing around or escalating unnecessarily
- Teams waiting for permission instead of taking ownership
- People avoiding challenge, conflict, or risk
- Customer needs being assumed rather than checked
- Collaboration looking polite, but delivery still feeling heavy
- “Meeting gravity” increasing while clarity drops
Step 2: Choose the behaviour
Next, we translate friction into a small number of observable behaviours.
Not values. Not slogans. Behaviours you can actually see.
Examples:
- “Make the decision with the people in the room.”
- “Escalate risks early, with options.”
- “Talk to customers before building.”
- “Close loops - every meeting ends with an owner and next step.”
Output: 1 to 3 priority behaviours to focus on first.
Step 3: Design a small experiment
We design a tiny change to test the behaviour in real work.
A good experiment is:
- Low-risk
- Easy to run
- Clear about who will do what, and when
- Measurable enough to spot signal
This is where culture change becomes practical.
Output: A simple experiment plan (what we’ll try, who’s involved, how we’ll measure, what “success” looks like).
Step 4: Run it and learn
We run the experiment and collect signals.
Signals can be:
- Behavioural: are people actually doing the thing?
- Qualitative: what are people noticing?
- Operational: did decision speed improve, did rework drop, did customer contact increase?
Then we ask:
- What changed?
- What got in the way?
- What surprised us?
- What should we adjust?
Output: A short learning summary and a decision: keep, adapt, or drop.
Step 5: Scale what works
If it works, we scale it carefully. Usually by making it easier to do the right thing.
Scaling might mean:
- A ritual (weekly customer check-in)
- A routine (decision log, pre-mortem, after-action review)
- A template (one-page experiment canvas)
- A leadership habit (visible role-modelling moments)
- A simple “how we do this here” standard
Output: A repeatable practice and a light rollout plan.
Step 6: Embed it
To make change stick, we embed it into the system so it survives busy periods and leadership changes.
Embedding points often include:
- Ways of working and team routines
- Role expectations and leadership rhythm
- Performance conversations and recognition
- Hiring and onboarding signals
- Decision-making forums and governance
- Tools, templates and meeting formats
Output: A small set of embed points and owners.
How we measure progress
(without turning it into a science project)
Culture measurement doesn’t need to be perfect. It needs to be useful.
I usually combine:
- Leading indicators: are the behaviours happening more often?
- Qualitative signals: what are people noticing and saying?
- Lightweight metrics: things like decision speed, cycle time, customer contact frequency, rework, handoffs, etc.
The goal isn’t a glossy dashboard. It’s learning fast, and building confidence in what works.
What you can expect from me
- Practical and direct, without being precious about theory
- Focused on behaviour and real work, not abstract “culture talk”
- Small experiments over big programmes
- Co-creation where it matters, clarity where it’s needed
- Honest feedback and a bias for action
- Low admin, high usefulness
A sensible starting point
Most work starts with a short discovery phase to spot the friction, choose the behaviours that matter most, and design the first experiments.
From there, we run the first loop, learn quickly, and decide what to scale and embed.
If you'd like to talk through your context, I’m happy to explore what a first set of experiments could look like.
Click the button below, or email me at matt@culturicity.uk