When we set out to co-design, we’re usually intending to get clear on a strategy or approach, find alignment across different groups and discover some new ideas together. If we’re putting together a formal workshop, we might plan, re-plan, create activities and facilitate the best we can. 

And often these types of things work well. But the expectations don’t always match reality. Ideas can get a bit lost in translation. Some participants invariably feel left out, or misunderstood. And worse, the designer might end up creating something altogether different. It’s like the workshop never happened. 

For some of those reasons, it’s natural for a designer to want to work in smaller groups, or by themselves, avoiding this way of working entirely. There’s less overhead and you can move faster if it’s just you and some focus time right? But less collaboration doesn’t work great either. Whenever I’m designing alone for too long or with limited input from other collaborators, I hit a dead end pretty quickly.

No one should design like this, even if it feels easier or more comfortable in the short term. I often think of this quote (now five years old) from Bob Baxley, reflecting on Apple’s review-heavy design culture: “If you ever found yourself sitting at your desk by yourself with your headphones on, stressing ’cause you felt like you had to figure it out on your own, something was really broken.”

Meet me halfway?

To make co-design more palatable, effective and integrated into our ways of working, we need something to fill that middle ground, somewhere between intensive multi-day workshops and flying solo. 

Recently, I found myself in a cross-functional group that had formed to help with product growth. We wanted to test out a lot of little experiments and see if any helped customers to get onboarded and set-up faster. This shared goal had given us valuable alignment to begin with, but we still needed to decide what experiments to run and how they should work. We had limited time and a big backlog of experiments to flesh out. How were we going to work together?

Once or twice a week, we’d come together to ‘frame’ or ‘shape’ an experiment. We’d discuss the problem it would solve, risks and a few possible approaches. It was never longer than an hour, and it was purposefully unstructured. There might be some ideation or diagraming or maybe we’d spend most the session on the right problem to solve.

Initially, I found this working session really frustrating. I’d think, shouldn’t the designer be doing this? How can we start to design without exactly understanding the problem? There wasn’t any time to optimise or improve the activities. I had the typical fear (of designers) that we were rushing into solutions, and we were! But we also needed to agree (or disagree) and commit, so we could move onto the next experiment and keep moving.

But slowly, I started to see some benefits. When it came time for me to actually design something, I was no longer starting with a blank slate. Many of the edge cases that would have me spinning wheels had already been resolved. And design reviews were more effective, since everyone had shared context and had already contributed significantly to the concept. I didn’t feel like I was shouting down a one-way telephone line.

Why did it work?

I think this sort of framing activity worked well for us because:

• Everyone involved was responsible for driving the same goal
• We’d already prioritized the rough idea, so we could focus purely on the how rather than the why
• Each experiment was scoped to no more than two weeks to build, so there was a natural expectation for designing to be similarly fast 

But most importantly, it succeeded because we allowed it to, without forcing it. If this was ‘co-design’, we never labelled it such. We need to allow co-designing to express in ways that are appropriate to your particular team or project. If you’re designing a new product from scratch, co-design will likely look very different. And that’s ok! It’s easy to get attached to how you think alignment and agreement should look. But if we’re too rigid in how we expect collaboration to happen, it probably won’t.