Help sort through generative or evaluative research data using an affinity diagram
Phases
Suggested Time
30-60 min per interview
Participants
Interviewers, researchers, observers, note-takers
Why do it?
We use the output of this activity to identify patterns in collected research data, improve our understanding of users, and identify action items we can take (including any re-testing to confirm any inconclusive findings). These findings can either validate or invalidate assumptions we’ve made about our persona (i.e. demographics, behaviors, concerns, motivators, etc). Output also supports conversations on the why of behaviors and what actions to take.
When to do it?
Ideally we do this after each user interview in order to analyze the information we captured during our user interview session while it’s still fresh in our minds.
What supplies are needed?
- Large Whiteboard or digital version like Miro
- Sticky notes with findings on them (stemming from research sessions), color-coded by research participant
- Dry erase markers
How to Use this Method
Sample Agenda & Prompts
-
Write the following categories on the whiteboard, taking care to leave ample space between them (to accommodate many sticky notes):
- Day in the Life
- Pain Points
- Goals
- Good to Know
- Magic Wand
- Actions
You can also create a Parking Lot at this point, though often we do this last (as we find sticky notes that aren’t actionable)
Tip: If you have already grouped your user interview questions into categories, use these categories in your synth instead for more targeted results. (i.e. Technology Usage, Systems & Integrations, etc.)
-
Place sticky note findings beneath the corresponding categories, including a section to identify the users (notes for whom should each have been written on different colored sticky notes)
Tip: Since several people may have taken notes, you will naturally find duplicates. When this happens, discuss as a team which sticky notes capture the comment best then discard the duplicates.
-
Once all have been placed, begin the process of grouping sticky notes. This can either be done as a team, pairs, or individually (depending on who is available at the time):
- Group sticky notes by a written sentiment from the point of view of the user
- Draw a smiley face that appropriately captures the sentiment (smiling, meh, frowning)
- Write the number of individual users that make up the group
- Circle each group when you are complete so you know that you don’t have to revisit the section
Remote Tip: Many video chat systems support “breakout rooms” or similar means to temporarily divide large groups into smaller groups for closer collaboration.
-
When all sticky notes have been grouped (or placed in the Parking Lot), do another pass to identify (a) actions to take and (b) findings to re-test
-
Take pictures of the output on the wall, ensuring you get a nice wide shot (which are great for sharing out or using as deck backgrounds) and clear, close-up shots (which are perfect for re-reading the “verbatims” that make up each sentiment group)
Success/Expected Outcomes
The goal is to walk away from research with actionable insights and an understanding of whether we’ve validated or invalidated any research objectives and to give design and/or product a set of actionable next steps to take
Facilitator Notes & Tips
As needed, build on the Persona data you’ve captured to-date. Take any new user insights and validate or invalidate the assumptions you have about your existing personas or create a new persona if you find new information that will support it.
Try using Trello (or something like it) to track your assumptions
It’s also helpful to share out findings at the subsequent iteration planning meeting (IPM, similar to SCRUM Sprint planning) so that the entire team – including anyone who wasn’t present during the actual research sessions – can benefit from what was learned.
You can alternately create your categories from the questions you are asking to your identified personas. These categories only tend to work for users of the same persona (i.e. do not mix user interviews with stakeholder interviews while synthesizing)
