skip to content

In UX design, what is an empathy map, and how does it differ from a persona in purpose and scope?

level: juniorimportance: nice to knowfreq 30%

answer

  1. a whiteboard, not a profile
  2. says, thinks, does, feels
  3. gaps between quadrants are the insight
  4. no time axis at all
  5. often a step toward personas

basics

~20 s

An empathy map is a quick, collaborative canvas capturing what a user or segment says, thinks, does and feels, used to align a team on research. A persona is a durable, goal-driven archetype the team designs against.

solid answer

~50 s

An empathy map is a lightweight canvas — commonly four quadrants for what a user **says**, **thinks**, **does** and **feels**, often with pains and gains — that a team fills in together from research notes. Its purpose is alignment and synthesis: it externalises what the team knows about one user or segment and exposes gaps, especially where what people say differs from what they do. A persona is a different artefact: a named, durable archetype with goals, behaviours, context and scenarios that the team designs against for months. So an empathy map is a snapshot of mindset, often one step on the way to a persona, while a persona is the reference cited in decisions. Neither shows steps over time; that is what a journey map adds. On a smart-home app, a map might show a user saying they love automation while overriding the heating schedule by hand every evening.

go deeper

for a junior

Recall the four common quadrants and that an empathy map is a quick shared snapshot of mindset, unlike a durable persona the team designs against.

for a middle

Explain why the gap between says and does is the insight, and how per-participant maps feed the synthesis of personas.

for a senior

Show when an empathy session is worth running — early alignment, contested pictures of the user — and how you keep inferred quadrants flagged as inference.

for a principal

Judge which lightweight artefacts a research practice standardises on, so teams align quickly without mistaking workshop output for evidence.

## What an empathy map is An **empathy map** is a collaborative canvas that captures a user's — or a segment's — mindset in one view. A team builds it together, usually in a short session, from material it already has: interview notes, field observations, support conversations. It is deliberately quick and rough. Its value lies less in the finished artefact than in the conversation it forces: people discover that they hold different pictures of the same user, and that some of what they “know” has no source at all. The common form puts the user at the centre with four quadrants around them, often with two extra areas for **pains** and **gains**. Variants exist — earlier versions split the canvas into what the user sees, hears, thinks and feels, and says and does, and some add a goal area — but the idea is the same: separate **observable behaviour** from **inferred inner state**. ## The quadrants and what to look for - **Says** — what the person said out loud, ideally verbatim quotes from research. - **Thinks** — what they seem to believe or worry about but may not say; this is inference and should be flagged as such. - **Does** — observed actions and behaviours. - **Feels** — emotional state: anxious, confident, irritated, reassured. - **Pains and gains**, where used — what frustrates them and what they hope to achieve. The insight usually lives in the **gaps between quadrants**. In a smart-home control app, a participant may *say* they love automation while the *does* quadrant shows them overriding the heating schedule by hand every evening, and *feels* records distrust: “I never know what it will do next.” That contradiction points at a design problem — the predictability and visibility of automations — that neither quadrant shows alone. ## How it differs from a persona | | Empathy map | Persona | Journey map | |---|---|---|---| | Core question | What is this user's mindset? | Who are we designing for? | What happens over time? | | Time axis | None — a snapshot | None — an archetype | Yes — stages in sequence | | Typical lifespan | Hours to a few weeks | Months, with revisions | Tied to one scenario | | Main use | Alignment, synthesis, spotting gaps | Reference for decisions | Locating pain points and opportunities | | Typical content | Quotes, observations, feelings | Goals, behaviours, context, scenarios | Stages, touchpoints, emotions, pain points | A persona is **durable and goal-driven**: a named archetype with goals, behaviours, constraints and scenarios, cited in decisions for months. An empathy map is **disposable and mindset-driven**: it usually carries little of a persona's goals, scenarios or context, and it often makes no claim to represent a whole segment. Many teams map each participant or each emerging segment, compare the maps, and then write personas from the patterns that repeat — so the map is frequently a **step toward** a persona rather than a substitute for one. ## When to use one 1. **Right after research sessions**, to get raw observations out of individual heads and onto a shared surface. 2. **When a team disagrees** about who the user is, to make the disagreement visible and ask which view has a source. 3. **Before writing personas**, to see which mindsets repeat across participants. 4. **To introduce an existing persona** to people new to the product, adding its felt experience to the facts. ## Common misuse - **Filling it from imagination**: a map built without research documents the team's assumptions, which is useful only if it is labelled that way. - **Treating inference as fact**: the thinks and feels quadrants are interpretations of what people said and did; mark them so. - **Averaging across very different users** into one map, which hides the contradictions the tool exists to reveal. - **Mistaking it for a journey**: an empathy map has no sequence, so it cannot show where in a flow a feeling arises. - **Keeping it as a deliverable**: its job is done once it has fed personas, a journey map or a design decision. The canvas is platform-neutral: the same session works whether the product lives on the web, on a phone, on a wall-mounted panel or in a device's companion app, because it describes the person rather than the interface.

  • Why does a gap between the says and does quadrants matter?
    People describe themselves through intentions and self-image, while behaviour shows what actually happens. When a smart-home user says they rely on automations but overrides them nightly, the gap exposes an unmet need — trust and predictability — that neither quadrant shows alone and that a self-reported answer would have missed.
  • Should a team make one empathy map per participant or one per segment?
    Both are used. Per-participant maps keep observations honest and make differences visible; comparing them afterwards shows which mindsets repeat and supports persona synthesis. A single map for a whole mixed user base averages away the contradictions the tool is meant to reveal.

saying these in an interview costs you the question

  • An empathy map is just a persona with its fields arranged in quadrants.
  • If the says and thinks quadrants disagree, the map was filled in wrongly.
  • Empathy maps come from team intuition, so they need no research input.
  • An empathy map shows the steps a user takes over time.