managing research output to create personas

This is an approach I took when managing research output to create personas for a recent project:

The Project:

• The research stage in redesigning the intranet for a  financial services firm
• I did not actually do the research!!! I was involved in a QA role working with a UE contractor
• The initial segmentation used to identify potential users was internal and based on roles and hierarchy

The Challenge:

Analyse research findings and reveal user groups and potential personas based on user goals, behaviours , attitudes, not the original segmentation. Basically my concern was that the original segmentation will drive the different types of user groups. So...

The Approach:

Forget the initial segmentation; it helped us identify potential users but did not necessarily reveal different behaviours. So, we took a step back and identified for each user interviewed, the following:

• Goals: What I want to achieve
• Motivations: What drives me in achieving those goals
• Behaviours: How is it that I respond to my role
• Tasks: What is it that i do on a dialy basis
• Issues / Pains / Frustrations I have with the current order of things
• Wishlist [for whatever we design / evaluating]
• General context of use including role, responsibilities, environment, etc.

The output was a sea of sticky notes for each category that we then split into different groups. The result was a set of brand new user groups with shared characteristics. Interestingly enough, the groups we identified did not map on to the original segmentation we used to recruit users. 

Now... off to designing.