Minibus
A weekly design practice session
run for my team at Canonical since 2024.
Minibus kickoff screenshot
I have spent more than twenty years designing across the whole spectrum, from visual to interaction to research to strategy, and the tools that came with that do not stop working when the problem is a team rather than a screen. I treat design team management like a design problem.
When I saw that there were shortcomings and needs in my org I did what I would do with any design problem. I understood the people, framed the need, designed a solution, measured it and iterated. Minibus is that solution. Over two years I have run 60+ sessions, rated every one of them, and changed the format when the quantitative and qualitative data told me to.
The research phase (My first year at Canonical)
I joined Canonical in mid 2023 to lead the enterprise design team, a group of around twelve cross-functional designers spread across EMEA with no clear forum for evolving. I did not arrive with Minibus in my pocket. The first twelve months went on the things that could not wait. The team had been without steady leadership for a while, so I spent that year putting out fires, building the levelling framework so people could see what growth looked like, and filling in as a designer myself where there was urgent need and nobody to meet it. That meant I understood the team and the products from the inside before I tried to change how they worked, and it meant the team saw me do the work before I asked them to do it differently. Raising the bar was always the intention and by the middle of 2024 I knew exactly where it needed raising:
The problems (What I found)
No shared framework to operate within as guidance. Craft conversations went round in circles because nobody had the same definition of good. Design review approved work it should not have, because designers presented their justification before reviewers had formed a view. No understanding of products as a suite nor what it meant to 'design for Enterprise'. And the team's development was almost entirely dependent on the individual attention I could give in one to ones, which doesn't scale.
The approach (What I decided)
The obvious lever was hiring. I had the standing to bring in senior people over the top of the existing team and let them set the bar. I chose not to because loyalty and a shared standard are built through investment and a mature critique culture cannot be recruited in, it has to be grown. So the mission was to raise the team I had, on strategy as well as craft, in a way that would compound rather than depend on me.
Minibus started on 19 June 2024 as a fortnightly Wednesday session. Every session is an exercise, a conversation, a workshop because the learning happens in the room and not as isolated homework. Every session is measured, because I wanted to know whether the team's time was being well spent and not assume it.
The solutions (What I did)
Minibus sessions have covered a broad range of design topics in order to address the problems. Ideation to find solutions. When something didn't land I redesigned the structure of the session rather than repeat the instruction. A few of the sessions that have done the most work.
- Communication: enabling designers to speak clearly about their work and ideas, developing from output to outcomes.
- Org design: a way for designers to locate their product in the broader context of the organisation and understanding its impact.
- What is good?: What makes a good designer and specifically at Canonical, which surfaced that the team held different definitions and let them see it for themselves.
- Calibrated omission: An exercise moving from observation to strategy by asking what a designer deliberately leaves out and why.
- Stakeholder as personas: Applying a method the team already trusts to the organisation around them.
- Conway's law: Reading the organisation through Conway's law, so designers could see why the products they own are shaped the way they are.
Metrics (What the team said)
Every session closes with a Return On Time Invested rating from one to five. I plot it against how often the sessions ran, because the question I cared about when the team asked to go weekly was whether doubling the cadence would dilute the value.
162 responses across 60+ sessions, from a team of around twelve. Two thirds of all ratings are five. ROTI measures whether people felt a session was worth their time, and it told me which sessions to keep and which to redesign. Whether the programme changed how the team designs is answered by the section below, not by this chart.
It did not dilute. Mean ROTI before the weekly cadence was 4.50 across 72 responses and 4.66 across 90 responses after it. The one dip, in the thin months of mid 2025, is the team's appetite outrunning the supply, and it is what prompted the move to weekly. The programme also held through the departure of the Global Head of Design at the end of April 2025, when it became the one fixed point in the team's week while the structure around it changed. So not only growth and evolution but stabilization too.
Outcomes (What the team did)
The measure that matters the most isnt whether the team was satisfied but what happens with them and in their products. Outcome over output.
- Enterprise pillars, first taught in Minibus, are now how designers organise a product roadmap, and the approach has formed a spec for product managers and lead engineers.
- UX+1 internal critique sessions have gained structure and objectivity.
- Reports received promotion based in part to responses based on communication and strategy.
- Ticket quality scores were 40% higher in the team doing Minibus sessions on tickets and their purpose.
Evidence (What the teams says)
Minibus has made the team understand their products deeper and in a more mature way.
The Minibus, where I've seen team growth in the last six months, thanks to the exercises and conversations you've facilitated.
Having the practical, actionable exercises reinforcing the takeaways really helps ground those big picture/meta topics into effective day-to-day usage.
It's helped me think a lot more strategically about my work, and how I can advance the role of design in my BAU.
Minibus helped me refine how I talk about content design within the Design Team and at Canonical.
Coming to Minibus has made me feel more connected to the team, and better able to drive for alignment across the board.
Challenges (What has not worked yet)
I have not yet brought the team to a shared definition of what makes a good designer. Three sessions have surfaced the same split (some designers treat visual craft and the more scientific parts of UX process as separate disciplines while I see them as one practice judged by the same outcome).This is important because of my ambitions to marry both in our solutions.
Next (Where it goes)
The most recent sessions have been run by senior designers rather than by me, choosing the topic, designing the exercise and holding the room. That is the direction I want it to go. A programme where senior designers work alongside me to drive the learning and outcomes of the department.
Takeaways (what this says about how I lead)
Minibus is a small thing that reflects how I lead more generally. I develop the people I have in conjunction with hiring great talent, because the team having a shared definition of great and the vision to understand their context onboards new talent more quickly. I could have mandated attendance and I never needed to, because the sessions are worth showing up for and the team asked for more of them. I treat strategy as a craft skill that every designer should hold, so the programme spends as much time on estimation, enterprise needs and judgement as it does on visual quality. And when I want behaviour to change I redesign the medium to make sure the message lands.
Back