Embracing Culture over Complexity: Lessons from Tokyo's Clean Streets
Every time I see a team reach for a new tool, a new process, or a new monitoring dashboard to fix a people problem, I think about Tokyo.
1 Nov 2023

Every time I see a team reach for a new tool, a new process, or a new monitoring dashboard to fix a people problem, I think about Tokyo.
The paradox
Tokyo has very few public trash bins. In most cities, that would mean litter everywhere. In Tokyo, the streets are spotless.
Why? Because people carry their own trash. They take personal responsibility for it. No elaborate collection systems. No enforcement campaigns. Just a deeply ingrained culture of ownership.
The engineering parallel
I've seen teams build complex systems to solve problems that culture would fix better. Elaborate approval workflows because people don't trust each other. Detailed time-tracking tools because managers don't trust their engineers. Mandatory code review checklists because nobody trusts the reviewers.
More process. More monitoring. More tools. And the root problem — lack of trust, lack of ownership — stays exactly where it was.
What I've learned
The most effective teams I've worked with had simple processes and strong culture. People owned their code. They cleaned up their own messes. They raised issues before being asked. Not because a system forced them to, but because that's just how things worked.
Building that culture is harder than buying a tool. It requires:
- Leaders who model the behavior they expect.
- Consistent reinforcement over months, not a single all-hands announcement.
- Trust — real trust — that people will do the right thing without surveillance.
The trade-off
Culture doesn't scale as easily as systems do. A strong culture in a team of eight doesn't automatically survive growing to eighty. You have to be intentional about preserving it as you grow.
But here's what I've seen: teams that invest in culture first and systems second build things that last. Teams that lead with systems and ignore culture build bureaucracies.
Sometimes the answer isn't a better bin. It's teaching people to carry their own trash.
Production-Ready Systems with LLMs and Agents
A live Maven cohort for engineers: eight 90-minute sessions where you build LLM and agent systems that survive real traffic, real cost and real failure.
Cohort 2 is running now. Join the waitlist for the next cohort.
View the live cohortKeep reading
- Your 2027 AI Plan Should Fit on One Page: Six Questions to Answer First
- The AI System Works. So Why Is Nobody Using It?
- If AI Had Existed When Your Workflow Was Designed, Would It Look Like This?
- AI Transformation Is Not Digital Transformation With a New Tool
- Your AI Pilot Saved Time. Why Didn't the Business Change?
- A Chatbot Answers. An Agent Does the Work. Here Is the Difference