#15 Culture by Design — Your Company Already Has One. Did You Choose It?

The poster on the wall says "meetings start and end on time." The team wrote it themselves, that morning, with some pride. Twenty minutes into the afternoon session, a member of the executive board drifts in, phone in hand, and drifts out again before the block ends.

I've watched a version of that scene more times than I can count. Nobody says anything. Nobody has to. Everyone in the room has just learned more about the real culture than the canvas will ever teach them.

Welcome to the new chapter of this newsletter. For fourteen editions we walked the Digital Leader Canvas — the mirror for the individual. As promised, we now pull back to the room the individual stands in: the culture of a team or an entire organisation. And for that I use another canvas, one I rate highly enough to build twelve-month change programmes on: the Culture Design Canvas by Gustavo Razzetti.

(New here? The Digital Leader Canvas — the individual-level companion to everything that follows — is still a free download: https://www.digital-leader-program.de/en/digital-leader-program-digital-leader-canvas)

One culture canvas among many — and why I use this one

First, an honest placement. There are plenty of culture frameworks and canvases out there; Razzetti's is one option among many, not the only door into this work. I keep coming back to it because it unites an unusual number of facets of culture on a single page. That makes it a superb visualisation tool: a large group of people can see, at a glance, the same picture of their culture — which matters most in bigger corporate transformations, where alignment is half the battle. And its limits belong on the table from the start: changing a culture takes far more than any canvas. What this one gives me is a home. The whole effort keeps returning to it, the concrete measures get derived from it, and the discipline of carrying them through has somewhere to report back to. Most of what culture work needs can be mapped there.

Minimize image

Edit image

Delete image

The one-pager itself (free at fearlessculture.design) puts Purpose and Values at the heart — literally, the two sit inside a heart shape — and arranges eight blocks around them: Priorities (your top three, written as even-over statements), Decision-Making (how is authority distributed?), Meetings, Norms & Rules, Behaviors (what do we reward, what do we punish?), Rituals, Feedback, and Psychological Safety.

You don't have to imagine what a filled-in canvas looks like: Razzetti has mapped well-known companies with it, and the examples are worth studying. Netflix — decisions made without a boss's approval, a culture of freedom and responsibility. Spotify — "ideas, not highest paid grade, win"; politics and micromanagement are what gets punished. Zappos — "wow our customers even over sales profit," complete with a pay-to-quit bonus for interns as a ritual. Airbnb — executive meeting notes shared with the whole company. Patagonia — managers trained to ask for feedback rather than give it. Even Wikipedia makes an appearance, running on one grown-up rule: "assume good faith." You'll find complete canvases for Netflix, Spotify, Zappos, Atlassian, Pixar and Slack on his site.

The deeper point sits underneath the tool: every organisation already has answers to all ten blocks. The only question is whether they were designed or just happened. Culture is like communication was two editions ago — you cannot not have one.

I've worked with this tool for almost three years now, in settings from a single team offsite to year-long culture programmes run top-down and bottom-up at once. One addition I always make: a block on good and bad failures — which mistakes do we want more of, because they mean someone experimented, and which must not happen twice? It sharpens the Psychological Safety conversation faster than anything else I've tried. (If you want to facilitate a session yourself, I've written up a six-step workshop guide — it's on the blog.)

But the canvas is the easy part. What I want to give you is what actually decides whether the exercise changes anything: three traps from practice, and where AI has started to help.

Trap one: the two-culture company

The scene from the opening is the most common failure mode I see, and it deserves its blunt name: many companies run two cultures. One for the top floor, one for everyone else. Meeting discipline is expected — except of the board. Feedback should be prepared with care and delivered on time — except downwards, where it somehow never arrives. The canvas says one thing; the corner office does another.

Employees resolve this contradiction instantly and always in the same direction: they believe the behaviour, not the poster. I made this point back in the Leadership Style edition about leadership principles, and it's doubly true for culture: it only moves if the most senior people visibly live it. A culture workshop where the executives exempt themselves isn't neutral — it's actively corrosive, because it teaches everyone that the written culture is fiction.

To be fair: different management roles carry different tasks, different pressures, and yes, sometimes different behaviours — a board member's week doesn't look like a team member's week. But "our role is different" cannot become the excuse for everything. The canvas doesn't demand identical days; it demands the same rules wherever the rules claim to be shared.

So if you take one rule into any culture work: the top level fills in the same canvas and submits to the same answers. No separate weather system on the top floor.

A quick aside for HR and people-development leaders: this is exactly the work I do with organisations — from a one-day culture offsite to twelve-month change programmes that run top-down and bottom-up in parallel. If your culture lives on a poster but not in the meetings, a short call is the place to start: we'll map out together what a realistic path looks like for your organisation. Book 20 minutes.

Trap two: the canvas is only as honest as the room

Here's the quieter problem. The canvas asks questions that need candour — what do we really punish? how safe is it to disagree? — but at the start of any culture process, the psychological safety to answer honestly usually doesn't exist yet. It's the chicken-and-egg of culture work: the thing you're trying to build is the thing you'd need to discuss it properly. So early canvases come out polite, cautious, and half-true. Nobody gives it to the room straight in front of their boss in week one.

This is where AI has earned a real place in my process — as an evidence collector that doesn't need courage.

Culture isn't what people say in a workshop; it's what happens between workshops. And much of that leaves traces that can be read at team level, with consent, without pointing at anyone: what the meeting load really looks like, how decisions actually travel, what anonymous survey answers say when they're synthesised properly rather than skimmed. An AI can aggregate all of it into a current-state hypothesis of the canvas — a draft of how this organisation demonstrably behaves, not how it hopes it behaves.

That changes the opening question of the workshop. Instead of "what's our culture?" — which invites poetry — you start with "here's what the evidence suggests; where is it wrong?" Disagreeing with a document is far safer than criticising a boss, and the honesty arrives a full workshop earlier. (The guardrails from the Digital Skills editions apply unchanged: team level, never the individual; consent, not surveillance. Measure culture covertly and you've destroyed the safety you were measuring.)

Two smaller AI roles round it out. In the room, it's a tireless consistency checker — "your Values say courage, your Behaviors punish failure; which is it?" — and merciless with weak even-over statements. And after the work, it becomes the drift-watcher: a scheduled agent (the same pattern as the KPI agent from Digital Skills, Part 2) that quietly checks whether the new meeting norms and feedback rhythms are actually being lived, and flags it early when the canvas starts becoming a poster.

Trap three: culture as a project with an end date

The third trap is built into how companies buy change: as a project. Twelve months, sometimes less, then "transformation complete."

A canvas session produces more change material than any organisation can absorb quickly — better feedback processes, new meeting rules, the discipline to finally use the project tools properly. Try to implement all of it inside a year and the everyday business pushes back, momentum dies, and the whole effort gets remembered as "that culture thing we did." I've come to believe the framing itself is wrong: culture isn't a project, it's a practice. The environment shifts, technology shifts, people leave and arrive, new generations walk in with different expectations — a culture that was "finished" last year is already out of date. The organisations that get this right work on culture the way this newsletter keeps preaching for individuals: continuously, in small increments, year after year. The 1% method, at company scale.

There's a number worth recalling here from the Energy & Focus edition: global employee engagement sits at just 21% (Gallup, 2024). That's what culture-left-to-chance looks like at scale. It didn't get there in a year, and it won't be fixed in one.

And inside this trap sits a familiar face from the Digital Skills editions: the readiness gap. Whether an organisation can work on culture with technology — evidence collection, drift agents, honest data — often splits along the same line as everything else: younger employees arrive energetic and full of ideas, while some of the senior people who hold the decision power don't see why any of it matters, perhaps because seeing it would mean changing habits of their own. When the people most protected by the status quo are the ones approving the changes to it, culture work slows to the speed of their comfort. Naming that dynamic — respectfully, but out loud — is often the first honest act of the whole process.

Your turn

You don't need a workshop to start. Pick the smallest honest test:

Take one block — Meetings is the easiest — and write two answers side by side: what our culture says, and what actually happened in the last two weeks. If the two match, take the next block. If they don't, you've found your first piece of real culture work — and it costs nothing but candour.

If you lead: do the same exercise for your own behaviour first. The two-culture trap starts wherever the person with the most power exempts themselves.

And if you want the machine's help: ask an AI to draft the current-state canvas from what you can share safely — meeting statistics, anonymised survey text — and argue with the draft as a team. Disagreeing with a document is the cheapest psychological safety hack there is.

Next time: the thought experiment this whole chapter has been building towards — if you built an organisation from scratch today, AI-native, how would each of these ten blocks change? My short answer: culture stops being decoration and becomes specification. The long answer is the next edition.

And before you go — this is the first edition beyond the canvas walk, so I mean this more than usual: tell me whether this new direction works for you. One sentence in the comments or by reply is plenty. I read every one.

See you then.

The Digital Leader Canvas is free for individual use. And if you're in HR or L&D and want to bring these themes into your organisation, here's that link again — book a 20-minute call. You can also explore the Leadership Essentials Workshop Series for teams.