My team at IBM is working on a project that could potentially affect a large span of organizations. At the moment, we are just wrapping up the beginning user research phase. It has been a challenge, though, because our users do not fit one set type. Instead, they all have different titles and roles.
Finding a way to unify them into well-rounded personas has taken its toll.
But, after conducting roughly 20 user interviews and a handful of ethnographic inquiries, our team started developing personas. We tried the traditional format:
53 years old
Background in ______
One wife, two kids
But, that simply wasn’t working. Due to varied titles and roles, what should have been three personas:
Was quickly becoming thirty-three personas, with a mix of:
you get the idea.
Title vs. Task
Over lunch, our team decided that we would have to consolidate. Although our users appeared to be vastly different by title and role, they had some things in common:
The tasks they needed to complete were all largely the same. So, we quickly shifted our view from extremely individualized role-based personas to task-based analysis.
For this, we developed target groups.
Target Group A
Task 1, 2, 3
Manager X-1, Manager X-2, Manager X-3…etc, etc
This allowed us to identify our users as three unified groups, instead of the fragmentation that was quickly occurring with narrower personas.
In the end, we still used traditional personas too.
But, we used them as representatives of a much larger group of people.