In defence of not doing the research first
Analysis paralysis is a legitimate danger. We need to trust our own design expertise and experience more. Until we create, we can't really learn.
We know that research is an important part of the product design process. It gives us insights into who our users are, and the kind of problems that they face. We use it to guide our decisions, and shape what it is we deliver. But there are times when we lean too far into that research. We second guess ourselves, not trusting our own skills and experience to give us a starting point to start building. We could move faster and start learning more valuable lessons if we didn’t bog ourselves down with research too early.
When we were preparing for an initial release of the the Salt Design System at J.P. Morgan, we knew we’d have a simple website to go with the first few components. But for some reason we were moving really slow at delivering designs and information architecture for that site.
Turns out that this was something that I hadn’t learned about J.P. Morgan’s culture. The company has a really robust research capacity, including very talented researchers on our team. They were going through a very rigorous research process about what should go into the design system site. Interviewing users, determining problems, journey mapping, defining personas, etc.
The thing was, this was a team that had been working in the design systems space for several years. The Salt Design System was the next evolution of a legacy design system which the corporate investment bank had been using for several years. A legacy design system, and a legacy design system website, that the team had been running. This was a team including members who spoke about design systems, attended design system events, had extensive domain knowledge.
It wasn’t the right time for that kind of exploratory research. We knew how to design and build a simple design system website. We were being too rigid in following a research process that didn’t account for pre-existing knowledge.
I’m a data driven design and product leader. I’m entirely bought into the value of research. I want data that takes the subjectivity out of design opinion. But there’s also an irony in how often I see design system teams - building an offering that aims to reduce replication of effort - try and reinvent the wheel about the design system itself! We can, and should, trust our expertise and experience to develop a hypothesis of what will work, build it, and then measure and research the validity of that hypothesis.
We’ll learn far more by building than merely by thinking about what to build. I want to build so that the results of our research has more value. In the case of the Salt Design System, that site has morphed and evolved after its initial release; as every offering will if its being responsive to its users. But there needs to be a starting point to evolve from, otherwise we’re forever frozen waiting to build a perfect solution.