What needs to be finished first
At least 5 interview sessions or comparable observations with the target user group are available as notes and quotes.
The user type to be mapped (persona, segment, stakeholder group) is defined clearly so everyone in the workshop has the same subject in mind.
What needs to be ready before start
Empathy map template (classic 4 or 6 quadrants: Says, Thinks, Does, Feels; optional Sees, Hears); whiteboard or digital board; one color for evidence stickies, another color for assumptions; research highlights ready to reference.
A facilitator; 4-8 team members from product, design, engineering, and optionally stakeholders; ideally at least one researcher with direct interview contact.
Research notes with quotes; user type description; workshop purpose (kickoff empathy vs assumption check vs synthesis); current hypotheses about the user.
30-60 min
Place template large on wall or board. Fix user type at the top. Define two sticky colors (for example blue = proven evidence, yellow = assumption). Set timer to 45 min. Rule: one concrete statement per sticky.
The one question this method answers
What does our user say, think, do, and feel, how do we know this, and where are our blind spots?
Marker: Phase
| Step | Duration | Action | Hint |
|---|---|---|---|
1Phase 1: Activate user type | 5 min | Name the user type aloud with a short description. If there are multiple personas: one map per user type only. Clarify purpose of the map (shared picture, assumption check). | If the team does not share a persona or user type, the map drifts immediately. Set anchor first, then map. |
2Phase 2: Fill says and does | 10-15 min | Add quotes from research notes as Says stickies (blue). Add observed behavior as Does (blue). If evidence is missing, mark Does as assumption (yellow). | Says and Does are the observed base and stronger foundation. Fewer evidenced stickies are better than many invented ones. |
3Phase 3: Add thinks and feels | 10-15 min | Thinks and Feels are interpreted, not directly observed. For each sticky add justification from a quote or observation. If no justification exists, mark as assumption and investigate. | Thinks and Feels are inference fields. If all stickies there are assumption-based, the map is more hypothesis than evidence. |
4Phase 4: Patterns and contradictions | 10-15 min | Review the map as a whole: are there contradictions (says X, does Y)? Are there clusters of statements? What is surprising? Mark these points as insights. | Contradictions between says and does are often the strongest insights. If no contradictions appear, depth or source volume may be insufficient. |
5Phase 5: Assumptions and next actions | 5-10 min | Prioritize yellow assumption stickies: which are critical, which need validation? For each top assumption define a validation step (interview question, survey, A/B test) with owner. | An empathy map without an assumptions backlog is a knowledge snapshot. With assumptions backlog it becomes a research driver. |
What comes out at the end
Empathy map as an image or digital export with user type label, 4-6 quadrants, clearly marked evidence and assumption stickies, identified insights, and assumptions backlog with validation plan.
One map per user type with date. For larger research rounds create a new version and keep older ones as comparison baseline. Track assumption status (open/validated/rejected) in assumptions backlog.
Compact working template for Empathy Mapping with context, input, output artifacts, and next step.
What is this method used for?
Which question should be answered at the end?
Which data, observations, or materials are available?
Owner, date, success signal.
Concrete filled scenario, fictional example
Concrete filled scenario, fictional example
Recognize symptoms and steer against them
All stickies have one color, nobody can identify what is evidenced.
Keep two colors strict: evidence (with quote source) vs assumption. Make it immediately clear while reading what is knowledge and what is hypothesis.
Map mixes statements from different user groups; insights become unusable.
One map equals one user type. For multiple types, create multiple maps. Otherwise you end up with an average with no signal.
Quadrants are full of interpretive statements without rationale.
Only add thinks and feels when they are derivable from says or does. If only gut feeling, mark as assumption and update assumptions backlog.
Team guesses about users and cannot validate statements with interviews.
At least one researcher with direct interview contact in the room. If not possible, postpone the workshop or mark it as assumption-only output (yellow only).
Map is created and stored in wiki, but assumptions are not validated.
Make assumptions backlog with owner and date mandatory at workshop end. Otherwise the map is just activity.
Done signals checkable in under a minute
Go to the profile for purpose, similar methods, and sources or continue to the next method in the catalog.