A typical project for Third & Grove follows a proven process: discovery and research leads to strategy, strategy leads to prototypes, and prototypes to testing. From there we move on to development once we have a good handle on the best approach. There are shifts here and there, but for the most part we’ve got it dialed, and it works.
A few months ago, however, we worked on a project as UX and design support only. If we were playing buzzword bingo “Tactical” would be worn out on our boards—that type of thing. We were partnering with a client who was vocal about the fact that the traditional user research approach (personas, experience maps, data, more data) would just slow down the process of launching new features.
At the beginning, I didn’t get it. I asked repeatedly, almost drone-like, about how they wanted us to do the strategic work to support the design and was unceremoniously shut down each time. In hindsight, I must have seemed desperate—they were essentially cutting me out of the project.
Our team had reservations too. It felt a bit awkward not only because we were used to a certain project rhythm, but because, frankly, it was more challenging. They wanted ideas, concepts, and prototypes based on our previous work, our technical knowledge of their site, and our collective understanding of their customers. Not quite a 5 day sprint, but not far off either. Call it strategy-lite, or lean—turns out they weren’t cutting out strategy altogether, they were instead pushing us to work differently.
Here’s the thing: it worked really well. After finding our mojo, we were pushing high quality work in days and weeks—a similar project with the standard process would still have been in meetings by the time we were launching big features.
As someone who has Strategy in their job title, this has been a bit of a sticky wicket. On one hand, the gospel of The Redesign tells us that research is good. That interviews and personas lead to empathetic (read: better) design. That a sexy presentation with bold statements and animated experience maps is what you do. I even wrote a 2,000 word post about how important data is to the design process—and I stand by it.
But boy, does it all take time. If everyone has their act together, you’re looking at a minimum of two weeks, sometimes four…often six or eight if schedules are prickly. And while there is unimpeachable value to the research work, it sometimes feels like a formality—a nice to have before everyone can get started.
When we do get started, we’re invariably building a prototype that is informed by the research, but by no means dependent on it. As others have found, some of the best data comes not from the upfront research, but from testing the prototypes (hallway testing, interviews, eye tracking, a/b with live traffic—whatever you’re into). Why not skip the step when you can afford it?
And yet, building a bespoke approach for every project and every client is impossible. We (agencies) need process because it helps us develop discipline, helps us train new team members, helps us learn from mistakes and improve on the next go-round. So what’s a strategist to do?
I don’t have a perfect answer, but we’re trying different things on. In-depth research moves closer to non-negotiable in my mind if one of three variables is present: if a full rebranding effort is happening, if there is a technical challenge, or if a completely new product is launching. In that case, I still see plenty of value in a slower, more thoughtful approach. But if a client just wants/needs a refresh? If the timeline is tight? If the agency before us did persona work? If we all (client + TAG) have a good gut feeling about where to go based on our (often under-utilized) collective experience? Then we’re shrinking the cycle, cutting out meetings, cutting out a lot of research, and moving quickly to prototypes so we can get live feedback first.
Right now our projects are probably 70/30, but I expect that smaller chunk of rapid prototyping to continue to grow as the tech enables us to move and test quickly. So yeah, we're processing....