Indi Young published her book “Mental Models: Aligning Design Strategy with Human Behavior” in 2008. I had never heard of it until 2018 and a co-worker at Red Hat suggested it as a research method for a new project.
This method, which produces a “mental model” at the end, is good for capturing what users are doing, and comparing that against what your solution does for them. Young defines it in the book as “Mental models give you a deep understanding of people’s motivations and thought-processes, along with the emotional and philosophical landscape in which they operate.” Instead of just looking at tasks or tools, or even defined processes and flows, the mental model captures a wide swath of non-linear data about users and how they relate to the service you’re offering.
The resulting model is rich and detailed, capturing things that wouldn’t be captured in a linear process. The interview process is open and loose, full of discovery. The model is theoretically good for years, as it focuses on human behaviors – which don’t change fast. This is an exciting way to do research.
For a result that will last years, you have to expect that it will take time to develop. And oh wow did it take time to do. I conducted about a dozen hour long interviews, then transcribed them, then combed them, then grouped the tasks, then grouped the tasks again, then presented them. The read-out alone took 3 hours! And that was rushed! This is a serious research activity, and it must be budgeted for as such.
Is it worth doing?
After reading the book, and working through the method, I have an inkling of what projects this is good for. This method is good for:
- Green field/blue sky ideas
- Services that have a variety of reasons for people to do them (in the book, Young uses “going to the movies” as a frequent example. It works very well.)
- Processes that have many linear paths and possible branches and choices
- Teams that are interested and bought into the general idea of UX research
Turns out, that it wasn’t a good match for my particular project. The project I was working on:
- Was always done in the same way every time (albeit highly inefficiently)
- Had a team that was new to UX research
- Had a researcher (me!) that wasn’t strong enough to push though a non-ideal method.
However, it wasn’t a complete wash! The resulting model is still valid, and the project is ongoing. Perhaps it will become more useful as the project gets out of a beta phase, and into a place where we are looking for more features and functionality to serve users.
Plus, I picked up some skills on research tools, and even built a generator that will build these models automatically. Look out for a post about these soon.