Method 26 of 100 · Stage 3: Talk to your future customers

Empathy Map

Sort interview quotes into says, thinks, does and feels, then hunt for the gaps between what customers say and what they actually do.

What is an empathy map?

An empathy map is a one-page board split into 4 boxes, says, thinks, does and feels, that you fill with evidence from customer interviews and observation. It forces you to look at one type of customer from the inside, and it makes one thing visible fast: the places where what people SAY and what they DO don’t match.

The map was born at XPLANE, the visual-thinking consultancy founded by Dave Gray, and spread through the book Gamestorming (XPLANE’s own history of the tool). Later versions add a goal at the top plus “sees” and “hears” boxes. For a first-time founder, the 4-box version is plenty.

When should you use it?

  • Right after 5 to 10 problem interviews, while the conversations are fresh.
  • When your notes are full of facts but you can’t explain WHY customers behave the way they do.
  • Before writing your landing page, so the words match how customers feel and talk.
  • For any business type. Consumer products show strong feelings. B2B maps are just as useful, because work fears (looking bad to the boss, losing a client) shape every purchase.

Skip it if you have no interviews yet. An empathy map filled from imagination is a mirror of your own assumptions, and it will look convincing, which makes it dangerous.

How do you do it, step by step?

  1. Name one customer type and one situation. Write at the top: “Parents of a child with a food allergy, planning the weekly shop.” A map for “all our users” ends up vague in every box.

  2. Draw the 4 boxes and set the rules for each.

    • Says: exact quotes only, in italics or quote marks, with the interview code (I3).
    • Does: actions they described step by step or you saw: routines, tools, workarounds, money spent.
    • Thinks: what they believe or worry about but rarely say out loud. Mark these as your interpretation and add the clue that led you there.
    • Feels: emotions with evidence: a sigh, a laugh, words like “dread”, “relief”, “embarrassed”.
  3. Fill it from your notes, one sticky per item. Go through each interview and write 5 to 15 stickies. Use one colour per interviewee so you can see later whether a box is driven by one loud voice or by many people.

  4. Cluster duplicates inside each box. Stack stickies that say the same thing and write the count on top: “Reads every label in the shop (6 of 8).” Counts turn a pile of quotes into evidence.

  5. Hunt for contradictions between says and does. Draw a red line between any pair that clashes: “I’m very organised” next to “Buys the same 6 foods every week because checking new ones takes too long.” Write each contradiction as a sentence below the map. These sentences are the most useful output of the whole exercise.

  6. Write 3 needs in the form “needs a way to…, because…”. Base each one on a contradiction or a strong feeling: “Needs a way to try new foods without re-checking every label, because the fear of a mistake outweighs the boredom.” These needs feed your offer and your concierge test.

  7. Check the thinks and feels boxes with 2 people. In your next interviews, test your interpretations gently: “Some parents told me they feel alone at birthday parties. Does that match your experience?” Keep what is confirmed. Cross out what isn’t.

What does a worked example look like?

Say Julia, a hypothetical founder, wants to build a meal-planning service for parents of children with food allergies. After 8 interviews, her map looks like this:

Box What she found
Says “I’m good at planning, I’ve got it under control” (I1, I4, I6); “I’d love more variety” (5 of 8)
Does Buys the same 6 to 8 safe products every week (6 of 8); spends 40+ minutes per shop reading labels (5 of 8); keeps a photo album of “safe” packaging on her phone (3 of 8)
Thinks “If I get one label wrong, my child ends up in hospital” (inferred from I2 and I7, both described past reactions minute by minute)
Feels Tired, on guard, lonely at birthday parties and school events (sighs and long pauses in 5 interviews)

The contradiction jumps out. Parents SAY they want variety, and they DO the opposite, because every new product means a new risk. What would sell is trust: a list of checked products, updated when recipes change.

Julia’s first offer changes from “100 allergy-friendly recipes” to a weekly list of 10 checked, new-to-you products for one allergy. That is a much smaller promise, and every line of it comes from the map.

Which free tools help?

Tool What you use it for Free plan
Miro Online board with empathy map templates and coloured stickies Free plan with a limited number of editable boards
FigJam Whiteboard by Figma for mapping with a co-founder Free plan with a limited number of files
Google Slides A simple 4-box slide when you work alone Free
Paper sticky notes The fastest version on a wall or table A few euros

What mistakes should you avoid?

  • Filling the thinks box with your hopes. “Thinks our app would save time” is your pitch dressed up as their thought. Every “thinks” sticky needs a clue from the interview.
  • Paraphrasing in the says box. Your summary loses the words customers use, and those words are your future headlines. Copy quotes exactly.
  • One map for everyone. Mixing two segments averages out the feelings. Make a separate map for each.
  • Stopping at a pretty board. The map is only done when you have written the contradictions and the needs underneath.
  • Letting one interview dominate. If all stickies in a box share one colour, that box is one person’s view.

How do you know it worked?

You can name at least 2 clear says/does contradictions, each backed by 3 or more people, and you have 3 written needs that you could not have written before the interviews. Our rule of thumb: if every need on your list matches what you believed on day 1, the map has just confirmed your bias. Look again, or interview people who disagree with you.

A strong sign: someone who reads your map, without hearing your idea, suggests a solution close to yours.

What should you do next?

Fold the map into a customer persona built from real data, and if the “does” box is thin, watch people at work with contextual inquiry. Then merge all your sources in interview synthesis. If you’d like an AI co-founder to question your interpretations, Fe/male Switch builds that into its startup game for women founders.

FAQ

What are the 4 quadrants of an empathy map?

Says, thinks, does and feels. Says holds exact quotes, does holds actions and habits, thinks holds beliefs and worries they keep to themselves, and feels holds emotions. Some versions add sees, hears and a goal.

What is the difference between an empathy map and a persona?

A persona describes who the customer is: situation, trigger, budget, channels. An empathy map shows what goes on inside them in one situation, and it is the better tool for finding hidden needs. Make the map first and use it to fill the persona.

Can I make an empathy map without interviews?

You can sketch one as a list of assumptions, but label it that way. Its real power comes from quotes and observed behaviour, so treat a map without interviews as a plan for what to ask.

Updated 26 October 2026 · Written by the Fe/male Switch team

Methods that go well with this one

Method 21

Problem Interviews

Ask people about the last time they hit the problem, never about your idea, to learn whether the problem is real and costly.

Customer discoveryBeginner45 minutes per interview, 2 to 3 weeks for 10 to 15
Method 25

Customer Persona from Real Data

A one-page profile of your first customer where every line traces back to an interview quote or a survey number.

Customer discoveryBeginner2 to 3 hours after your interviews
Method 29

Contextual Inquiry (Shadowing)

Watch customers do the real task where they normally do it, and write down every workaround, spreadsheet and sticky note you see.

Customer discoveryIntermediate60 to 90 minutes per session, 4 to 6 sessions
Method 30

Interview Synthesis

Turn a pile of interview notes into counted patterns with affinity mapping, then make a clear go, pivot or stop decision.

Customer discoveryIntermediate3 to 5 hours for 10 to 15 interviews
Method 33

Concierge Test

Deliver the service by hand to your first 5 to 10 paying customers and log every step before you automate anything.

No-code prototypeBeginner2 to 4 weeks

Go deeper