What is Empathy Mapping?

An empathy map is a canvas, usually divided into four quadrants around a central user or segment, that captures what the user says (direct quotes and statements from research), thinks (what occupies their mind, including what they wouldn't say aloud), does (observed actions and behaviors), and feels (emotional states, worries, and hopes), with many versions adding sections for pains and gains. The format was popularized by Dave Gray and colleagues as a lightweight team exercise, and it has become a standard step in design thinking workshops and product kickoffs. Its value is synthesis: a team that has done interviews, sessions, and field visits fills the quadrants from that evidence and, in doing so, notices the gaps (we know what they do and have no idea what they feel) and the tensions (they say the process is fine and their behavior shows they avoid it) that drive design decisions.

Building One From Evidence

1. Choose the user.
One person or one tight segment per map; a map of "our users" is a map of nobody. Where personas exist, a map per persona; where they don't, a map per real interviewee is a good way to start.

2. Fill from research, note by note, with sources.
Each sticky note is an observation from a specific interview, session, or recording, tagged with its participant, exactly as in affinity mapping. Quotes go in "says", observed actions in "does", and inferences about thinking and feeling are labeled as inferences and grounded in something the participant said or did.

3. Do it as a team, silently first.
Everyone who touched the research adds notes independently before discussion, so the map reflects the evidence rather than the most confident colleague.

4. Look for the gaps and the contradictions.
An empty "feels" quadrant is a research plan. A "says" that contradicts a "does" is the finding: the say/do gap, made visible.

5. Extract needs and open questions.
The map's output is a short list of needs, pains, and things the team still doesn't know, which feed jobs, opportunities, and the next round of research.

Where It Fits

Empathy maps are narrower than journey maps (one user's state, not a sequence of stages) and lighter than personas (a working synthesis, not a stable archetype). They sit naturally between research and design: after a round of interviews or recorded sessions and before ideation, as the exercise that gets a whole team to internalize what the participants said and did. Their speed is the point, and their speed is also their risk.

The Trap: Empathy Without Evidence

The common failure is the workshop that fills all four quadrants from imagination: a team that has never spoken to a user writes what they assume users think and feel, and the confident completeness of the map hides the fact that it is fiction. Empathy that wasn't earned through research is projection with sticky notes, and it is worse than an empty map because it feels like understanding. The test for every note is the same as for any research claim: which participant, which session, what did they say or do? A quick recorded study (a handful of screened users answering on video, the kind a Ballpark study fields in a day) is cheap enough that no map needs to be filled from guesswork.

What to Remember

Empathy mapping organizes research evidence about one user into says, thinks, does, and feels, so a team can see the person whole and notice what it doesn't know. Choose one user, fill from sourced observations, work silently before discussing, mine the gaps and contradictions, and extract needs. Fill it from research or don't fill it; a map of assumptions is the most convincing way to be wrong about people.