7627RE: [agile-usability] Building Agile Personas
- Apr 11, 2013
You are correct about personas. For agile design and development, there are better approaches. The first is to carry out model-driven inquiry instead of an open ended ethnographic field study or extended contextual inquiry that generates tons of “data”. Instead, you start with exploratory modeling to identify areas for short, focused field inquiry. Then the model is refined and completed as needed for the immediate design.
The second is to employ user roles instead of fully fledged personas. The latter include a great deal of “noise” that serves primarily to make the persona seem real but has little or nothing to do with good design. User roles, as constructed in usage-centered design or activity-directed design, are collections of responsibilities, needs, interests, and expectations of the actor within the role within an activity. A role focuses on the relationship between an actor in role and a designed artifact of interest. In other words, a user role is centered on the important stuff with the greatest impact on good interaction design and user experience in interaction with the designed artifact. It does not waste time on all that “other stuff” that makes personas fun and interesting but offers little leverage for the agile designer.
For agile projects: model-driven inquiry with user roles instead of ethnographic/contextual inquiry and personas.
Prof. Larry Constantine, IDSA, ACM Fellow
Universidade da Madeira | Madeira Interactive Technologies Institute
Gabinete 2.56 | int'l m: +44 792.447.5358
I am familiar with personas as Cooper describes in About Face, but am just getting started with Agile.
Building real research based personas is a time taking activity, and kind of waterfall-ish: research->analyze data->... ... ->publish personas.
For a sprint which is 2 or 3 weeks, how can we do all that without compromising on conducting qualitative research?
(I believe assumption personas are useless and sometimes counterproductive.)
Thanks in advance.
- << Previous post in topic Next post in topic >>