The O shaped designer
2026
We have spoken about different shaped designers for a few years. T shaped people who had a good broad base of skills and excelled in one area (adapted by Idean, now Frog into π shaped people, to describe people with two or more specialisms). Then M shaped designers, who had a better than broad solid base, and excelled in a number of areas of the craft. Then there were X shaped people - team leaders who could happily sit across numerous teams, viewing the work at a high level, able to get into the detail and terrific at joining the dots.
Now we see the advent of O shaped designers. Augmented by a raft of AI skills, a T shaped designer can expand into other areas. An excellent service designer can now pull together a working prototype of the digital component of the system or service they are designing in minutes. They can use a design system to ensure it is compliant, accessible and on brand. They can run a content design audit on what they produced. They can optimise the UX flows. They can build a working app, either using real data inputs or synthetic data housed in Supabase. They can plan and build a beta version, commit it to Github and push it live for testing. And they can design and deliver a meaningful test with real users. Will it be as good as if they were collaborating with the normal 15-20 other people such a project would have required 2 years ago? Hard to say. But would that project have been a success anyway? Also, hard to say. The difference is, you can get your answers faster, and with fewer handoffs and delays.
Is this better? It definitely has some advantages. Talented individuals can tackle large swathes of a problem space. They can go through a double diamond process in less time than an 8-person team could run a design sprint. They can maintain focus and take bigger risks, without other people forcing them to hedge and take a middle road.
And it has some risks. If the content design skill is weak, you’ll end up with a host of “it lands”, “earns its place”, “not x but Y” AI-isms and, worse, outright fabulations and hallucinations. The UX design skill will be pulling on what already exists, which might be fine, but will only be convincingly OK, sometimes not even that. The technology the beta is built on, without very comprehensive contexts and workflows, is going to be trash. All of which might be OK, depending on the context.
The one thing I think solves some of this is small teams of O-shaped experts. An engineer with deep technical expertise, who can design, plan, build and ship will spot the gaffs, gaps and issues immediately, will avoid them completely. A product owner with a deep understanding of the commercial realities of the problem space can design, build, plan and deliver, and will spot immediate product-market fit issues, unreal trade-offs and value erosion faster than any service designer. This was what was always envisioned in the idealised Three Legged Stool - three leads, bringing together their respective areas’ core skills and thrashing out the optimal solution.
The difference is, now, a team of three experts can do this work themselves. And you can structure it so the tension, human conflict and debate is additive, not reductive. You can bring these different mindsets and domain expertise together to deliver the next generation of services and products in a way that just wasn’t possible before. Although, anyone in a small to mid-sized agency would have told you that was exactly how they’d been doing things for years.