An empathy map puts the whole team inside the user's perspective. Four quadrants, one board. Use it to align on who you're designing for before you design anything.
Run an empathy map at the start of a design or product phase, after customer interviews (to synthesize what you heard), or when the team has diverging opinions about who the user actually is and what they care about.
The empathy map works best as a synthesis tool, you fill it from real customer data, not from assumptions. If you don't have customer research yet, use the session to surface your hypotheses, then treat every sticky as a question to validate.
Verbatim quotes from real customers, things they actually said in interviews, support tickets, reviews, or user tests. Put quotes directly here. 'I always lose the action items from our retros' is a Says sticky. 'Users find documentation unclear' is not, that's your interpretation.
What the user thinks but might not say aloud. Internal monologue, concerns, assumptions, beliefs. This quadrant requires inference, you're reading between the lines of what they said. 'I wonder if my team will actually follow through on this' is a Thinks sticky.
Observable behaviors and actions. What do they actually do, not what they say they do? 'Takes a photo of the whiteboard on their phone,' 'texts action items to the team,' 'opens a new Notion doc after every meeting.' These come from observation or analytics, not self-report.
The emotional state, positive and negative. Frustrations, anxieties, satisfactions, delights. Use emotion words: overwhelmed, skeptical, relieved, embarrassed. These often come from tone of voice in interviews, not explicit statements. One emotion per sticky.
Name the specific user and the specific situation you're mapping. 'Sarah, product manager, running a sprint retro with a five-person team.' Without context, the map is too abstract to be useful.
If you have interview recordings or notes, pull direct quotes and observations onto stickies. Don't sort yet, just get the raw material onto the board.
Place each sticky in the appropriate quadrant. The Says/Thinks and Does/Feels distinctions force precision, 'says' is verbatim, 'thinks' is inferred. That distinction reveals what you actually know vs. what you're assuming.
Look for contradictions between quadrants. What users say and what they do often diverge. A user who says 'we always capture action items' but does 'sends a photo to a group chat' is telling you something important.
Write 3 to 5 insight statements below the map: 'Users trust physical notes more than digital tools because...' Each insight should explain the tension or pattern, not just describe it.
Snap the board with BoardSnap. The AI reads all four quadrants and outputs a structured synthesis, the insight statements become action items for the next design iteration.
An empathy map built in a digital tool gets filled quickly and read rarely. One built on a physical whiteboard with the whole team arguing about which quadrant a sticky belongs in is a fundamentally different artifact. The argument is the value, it forces the team to be precise about what they actually know about the user.
Empathy maps are often the most insight-dense boards a product team produces. BoardSnap reads all four quadrants and preserves the raw content, Says quotes, Thinks inferences, Does observations, Feels emotions, in a structured format that can feed directly into design briefs and user story workshops.
A user persona describes who the user is, demographic, context, goals. An empathy map describes how they experience a specific situation, what they say, think, do, and feel. The persona is the stable profile; the empathy map is situational. Build the persona first, then use empathy maps to understand specific moments in that persona's journey.
Yes, but treat every sticky as a hypothesis. A hypothesis-based empathy map is useful for aligning the team before research and for generating interview questions. Just be honest about what's assumption and what's evidence, and plan to validate within two weeks.
Five to ten per quadrant is typical. More than fifteen and the quadrant becomes noise rather than signal. If you have too many, cluster and prioritize: which Says quotes are most representative? Which Feels are most intense?
The whole team. Engineering hearing customer quotes directly is different from reading a research report. The physical exercise of writing stickies and debating quadrant placement builds shared understanding that no written document can replicate.
After every major round of customer research. An empathy map built from three customer interviews is a starting point; one built from thirty is a foundation. Snap each version with BoardSnap, the evolution of the map tells the story of what you learned.
No exporting, no transcription. Snap the board, get the action plan.