Empathy Map: Guide, Practical Example & Template for Service Design

The Empathy Map is a visual tool that structures a user group’s perspective into four to six segments — what the person thinks and feels, sees, hears, says and does, plus their pains and gains. It was developed by Dave Gray at XPLANE and introduced to a wide audience through Alexander Osterwalder’s Business Model Generation (2010) [1][2].

What distinguishes the Empathy Map from other user research methods: it forces a team to reconstruct the inner state of a user beyond demographics and behavioral observations — thoughts, emotions, perceptions, influences. This bridges the gap between quantitative data (“42% of customers abandon the process”) and qualitative empathy (“Why does the customer feel overwhelmed at this moment?”).

Search for “Empathy Map” and you will find dozens of results showing the same Post-it poster. None distinguishes the original 2010 format from the revised 2017 format. None explains how the Empathy Map fits into a professional service design process — as input for Journey Maps and Service Blueprints. And none honestly identifies when you should use a persona, a user story format, or go directly to a Customer Journey Map instead.

This guide closes those gaps.

From XPLANE to Osterwalder: Where the method comes from

Dave Gray, founder of the consulting firm XPLANE and author of Gamestorming (2010), developed the Empathy Map as a workshop tool to help teams step into their target audience’s perspective [1]. The original format had six segments arranged around a central user: Think & Feel, See, Hear, Say & Do, Pains, Gains.

The real adoption came through Alexander Osterwalder, who included the Empathy Map as a companion tool in Business Model Generation (2010) — the book that also popularized the Business Model Canvas [2]. Osterwalder’s recommendation: before filling out a Value Proposition Canvas, create an Empathy Map for each customer segment. This made the Empathy Map a standard tool in the Lean Startup and Design Thinking community.

The 2017 revision

In 2017, Dave Gray fundamentally revised the format and published the updated version on Medium [3]. The key changes:

Aspect	Original (2010)	Revised (2017)
Segments	6 (Think & Feel, See, Hear, Say & Do, Pains, Gains)	4 quadrants (Says, Thinks, Does, Feels) + Goal
Starting point	No defined starting point	Begins with a clear Goal (user’s objective)
Pains & Gains	Integrated in the diagram	Removed — Gray argued they belong in a separate Value Proposition Canvas
Focus	Broad customer profile	Specific situation or task

Which version should you use?

The original version (with Pains & Gains) works best for early exploration phases when you want to develop a broad customer profile — e.g., at the start of a service design project, before creating personas. The revised version (4 quadrants + Goal) works best for specific situational analysis — e.g., “How does the customer experience the moment of filing an insurance claim?”

In practice, most service design teams today use a hybrid: the four quadrants (Says, Thinks, Does, Feels) plus Pains and Gains as separate lower fields [4].

Academic context

The Empathy Map has no dedicated academic foundational research — it is a practitioner tool, not an academic construct. Its theoretical underpinning lies in perspective-taking empathy, which Kouprie and Visser (2009) described as a central mechanism in user-centered design [5]. They distinguished four phases of empathic design work: Discovery, Immersion, Connection, and Detachment. The Empathy Map primarily operationalizes the Immersion and Connection phases — it helps the team step into the user’s perspective and build an emotional connection.

Gasparini (2015) investigated the Empathy Map as a pedagogical tool in software engineering and confirmed that it helps teams “gain insights into users’ feelings and needs” — but only when based on real research data [6].

The 6 segments in detail (hybrid format)

The format most commonly used in service design practice combines four quadrants with two lower fields:

1. Says

What does the user say verbatim? Direct quotes from interviews, support tickets, reviews. Not paraphrased, not interpreted — original voice.

Example (insurance customer filing a claim):

• “I don’t know which documents I need to submit.”
• “Last time, processing took three weeks.”
• “Can’t I just call someone?”

Common mistake: The team invents quotes that fit the desired outcome instead of using actual customer voices. Solution: Only use what was actually said in interviews, surveys, or support tickets.

2. Thinks

What does the user think but not say aloud? Thoughts, inner monologues, unspoken concerns. This segment requires interpretation — which is why it is both the most valuable and the most risky.

Example:

• “Will the insurance company try to avoid paying?”
• “I should have reported the damage immediately, not a week later.”
• “Hopefully I won’t have to explain this twice.”

Important: The boundary between Says and Thinks is the difference between what customers express publicly and what they fear privately. In interviews, you get both — if you ask the right questions and create a trusting atmosphere.

3. Does

What concrete actions does the user take? Observable behavior, not intentions. Best gathered through observation or behavioral analytics, not surveys (people notoriously misreport their own behavior).

Example:

• Searches the website for “report claim,” cannot find the link on the homepage
• Calls the hotline and is kept on hold for 12 minutes
• Photographs the damage with a smartphone but finds no upload function
• Asks a colleague who has filed a claim before

4. Feels

What emotions does the user experience? Fear, frustration, relief, confusion, anger. Emotions are the strongest indicators of pain points and delight potential.

Example:

• Uncertain: “Am I doing everything right?”
• Frustrated: “Why is this so complicated?”
• Suspicious: “Will the insurance company be fair?”
• Relieved (after confirmation): “Finally, someone knows about my case.”

Common mistake: Emotions are formulated too generically — “customer is dissatisfied” says nothing. Be specific: “Customer feels abandoned because no confirmation arrives after filing the claim.”

5. Pains

What obstacles, frustrations, and risks does the user experience? What stands in the way? What costs time, money, nerves?

Example:

• Uncertainty about processing status after filing
• Fear of financial burden if the insurance doesn’t pay
• Having to explain the same situation repeatedly to different contacts
• Feeling treated like a supplicant

6. Gains

What does the user wish for? What would be a positive outcome beyond expectations? Not just “claim gets paid” but also emotional and social gains.

Example:

• Immediate confirmation that the claim has been received
• Transparent processing status like package tracking
• The feeling of being taken seriously as a customer
• Proactive information about next steps

When is the Empathy Map the right tool?

Use the Empathy Map when:

• You want to understand a customer segment for the first time and don’t have personas yet
• You want to align a cross-functional team on a shared customer perspective — the Empathy Map as a workshop tool creates alignment
• You have conducted user research and want to visually synthesize the results before creating personas or journey maps
• You are facilitating a Design Thinking introduction and need a low-barrier tool that participants without design experience immediately understand

Use a different tool when:

Situation	Better alternative	Why
You need a permanent reference document for the customer segment	Persona	Personas are more detailed and longer-lived than Empathy Maps
You want to map the entire process a customer goes through	Customer Journey Map	Journey Maps are chronological and process-oriented; the Empathy Map is situational
You want to map frontstage and backstage of a service	Service Blueprint	Blueprints show the organizational side that an Empathy Map doesn’t capture
You want to understand stakeholder relationships and power dynamics	Stakeholder Map	The Empathy Map shows one person; the Stakeholder Map shows the relationship network
You have quantitative data and want to prioritize features	Kano model	Kano quantifies satisfaction impact; the Empathy Map is qualitative

Comparison: Empathy Map vs. Persona vs. User Story

Three tools that are often confused — but serve different purposes:

Dimension	Empathy Map	Persona	User Story
Purpose	Snapshot of user perspective in a situation	Permanent reference profile of a customer segment	Single requirement from the user’s perspective
Depth	6 segments, qualitative	Demographic, psychographic, goals, frustrations, quotes, scenario	1 sentence: “As a [role] I want [action] so that [benefit]“
Data source	Interviews, observation, research synthesis	Long-term user research, quantitative + qualitative data	Workshop, backlog planning
Lifespan	Workshop artifact (hours to days)	Months to years	Sprint cycle (weeks)
Team benefit	Build empathy, create alignment	Decision reference for product development	Formulate development task
Origin	Dave Gray / XPLANE [1]	Alan Cooper, The Inmates Are Running the Asylum (1999) [7]	Kent Beck / Extreme Programming (1990s)

The ideal sequence: Empathy Map (workshop, fast) → Persona (consolidated, permanent) → User Stories (derived from personas). Anyone who writes user stories without prior empathy work projects their own perspective instead of the user’s.

Step by step: Create an Empathy Map in 60 minutes

Preparation (before the workshop)

• Participants: 3–8 people from different functions (product management, customer service, UX, sales, engineering). The more diverse the perspectives, the more complete the map.
• Research foundation: At least 5 qualitative interviews or a substantial data basis (support tickets, NPS comments, usability tests). Without a data foundation, the Empathy Map becomes fiction — this is the most common and most damaging mistake (see “Common mistakes” section).
• Materials: Whiteboard or large paper (A1/A0), Post-its in 6 colors (one color per segment), markers, timer
• Prepare template: Pre-draw the 6 segments — a circle in the center for the user description (name, role, context), the four quadrants above (Says, Thinks, Does, Feels), the two fields below (Pains, Gains)

Phase 1: Define user and context (10 minutes)

Define who and in what situation the Empathy Map is being created for.

Too vague	Specific enough
”Our customer"	"Individual customer, 45, filing a water damage claim for the first time"
"The user"	"Claims processor in back office, handles 30 cases per day"
"New customer"	"Managing director, mid-sized company, evaluating insurance offers for 120 employees”

Facilitation tip: If you have multiple customer segments, create a separate Empathy Map for each segment. A map that describes “the customer” describes no one.

Write the user and their context in the center of the template.

Phase 2: Review research data (10 minutes)

Lay out the research foundation: interview transcripts, support ticket excerpts, survey results, observation notes. Each participant reads silently and marks relevant passages.

Facilitation tip: If no research data is available, do not run the workshop. Instead, plan a brief user research phase — 5 interviews of 30 minutes each are the minimum. An Empathy Map without data is a collective assumption collection masquerading as customer knowledge.

Phase 3: Fill segments (30 minutes)

Work through the segments one at a time — not all simultaneously. Recommended sequence:

1. Says (5 minutes): Easiest — direct quotes from interviews and support tickets. Each participant writes quotes on Post-its and places them in the quadrant.
2. Does (5 minutes): Observable behavior. What does the user actually do? What actions, what workarounds?
3. Thinks (5 minutes): Interpretation phase. What does the user think but not say? The helpful question: “If we could read this customer’s thoughts, what would we see?”
4. Feels (5 minutes): Name emotions — specific, not generic. Not “dissatisfied” but “anxious because no confirmation arrives after filing.”
5. Pains (5 minutes): Obstacles, frustrations, risks. What stands in the user’s way?
6. Gains (5 minutes): Wishes, hopes, positive surprises. What would exceed expectations?

Facilitation tip: During the Thinks and Feels phases, teams tend to project their own perspective. Ask the check question: “How do we know this? Is there an interview quote or observation that supports it?” If not, mark the Post-it as a hypothesis (e.g., with a question mark).

Phase 4: Identify patterns and prioritize (10 minutes)

Step back and look at the entire map:

• Identify clusters: What themes appear across multiple segments? If the user says “I don’t know what happens next,” does “call the hotline to check,” thinks “Hopefully the application isn’t lost,” and feels “uncertainty” — then “lack of process transparency” is a dominant theme.
• Uncover contradictions: Does the user say “I don’t care about price” but do “compare three quotes”? Such contradictions are the most valuable insights — they reveal where social desirability distorts self-reporting.
• Formulate top 3 insights: State the three most important findings as concrete assertions: “The customer has no idea what happens after filing the claim — and doesn’t dare ask because they fear being perceived as annoying.”

Example: Empathy Map for an insurance claims process

Context: A large DACH-region insurer is redesigning its claims process. The service design team has conducted 12 in-depth interviews with individual customers who filed a claim in the last 6 months. The Empathy Map summarizes the findings.

User: Maria, 52, individual customer. Filing a water damage claim for her apartment for the first time.

Segment	Contents
Says	”I didn’t know I had to report the damage within 48 hours.” / “The online form asked me things I can’t possibly know — like the exact amount of damage.” / “Nobody told me I’m not allowed to hire the contractor myself.”
Thinks	”Will the insurance try not to pay?” / “Should I have taken higher coverage?” / “Is my neighbor responsible — and do I have to prove it?”
Does	Searches “report water damage [insurer name]” on Google, lands on an FAQ page with no direct link to the form. / Calls the hotline and waits 8 minutes. / Photographs the damage — 14 photos from different angles. / Asks her sister, who “knows about these things.”
Feels	Overwhelmed: Too many forms, too much jargon. / Suspicious: “They probably don’t want to pay.” / Abandoned: After the online submission, no confirmation — 3 days of silence. / Relieved: Only when the assessor arrived on-site and said: “This will be covered.”
Pains	No understanding of the process or her own obligations. / Online form requests information she doesn’t have (damage amount, cause). / 3 days without feedback after submission. / Every new contact asks the same questions again.
Gains	Immediate confirmation with a personal contact and direct phone number. / Clear checklist: “Here are the 5 things you need, these are the next 3 steps.” / Status tracking like a package: “Step 2 of 5 completed.” / Proactive call: “Ms. M., we’ve received your report. The assessor will come on Thursday.”

Top 3 insights:

1. Information gap: The customer knows neither the process nor her obligations — and the insurer assumes both.
2. Communication gap: After filing, silence. The customer interprets silence as rejection.
3. Trust gap: The customer assumes the insurer is working against her — every delay reinforces this suspicion.

Next step: These three insights feed into the Customer Journey Map of the claims process. There, they are placed chronologically and linked to specific touchpoints. From the Gains, the team derives design principles — e.g., “Proactive communication within 2 hours of every status change.”

Note: This example is illustratively constructed to demonstrate the method in a service context. The insights are based on typical industry findings from the insurance sector.

The Empathy Map in the service design process: How it connects with other tools

The Empathy Map is not an isolated tool — its value comes from embedding it in the service design process:

Typical sequence:

1. User Research → Collect data (interviews, observation, surveys)
2. Empathy Map → Synthesize data and visualize the customer perspective
3. Personas → Consolidate permanent reference profiles from Empathy Maps
4. Customer Journey Map → Map the experience chronologically, identify touchpoints
5. Service Blueprint → Connect frontstage experience with backstage processes

The Empathy Map as bridge: It translates raw research data into a form the entire team understands — including stakeholders without a design background. A product manager who reads an interview transcript takes away different things than one who sees a structured Empathy Map. The structuring ensures nothing is overlooked: if the Feels quadrant is empty, the team hasn’t captured the emotional dimension — and knows it needs to investigate further.

Empathy Map: Template for immediate use

Use this template directly for your next workshop:

Preparation

• User and context defined (name, role, situation — no generic “customers”)
• Research foundation available (min. 5 interviews or substantial data basis)
• Template prepared (6 segments: Says, Thinks, Does, Feels, Pains, Gains)
• Participants from at least 3 different functions invited
• Post-its in 6 colors, markers, timer provided

Execution (60 minutes)

• Phase 1: User and context defined and written in the center (10 min.)
• Phase 2: Research data reviewed, relevant passages marked (10 min.)
• Phase 3: All 6 segments filled, 5 minutes per segment (30 min.)
• Phase 4: Clusters identified, contradictions uncovered, top 3 insights formulated (10 min.)

Quality check

• Every Post-it traces back to a concrete data source (interview, observation, support ticket)
• Hypotheses are marked as such (question mark)
• No segment is empty — empty segments indicate research gaps
• Top 3 insights are formulated as concrete, actionable statements

Next steps

• Empathy Map photographed and digitally documented
• Top 3 insights communicated to stakeholders
• Insights transferred into persona development or journey map
• Research gaps from empty segments or marked hypotheses documented as next research questions

5 common mistakes with the Empathy Map

1. Filling it without research data

Symptom: The team fills the Empathy Map in a workshop without having conducted interviews, read support tickets, or observed users beforehand. The map is based on “what we think the customer thinks.”

Why it hurts: An Empathy Map without a data foundation is a collective assumption collection. It validates the team’s existing biases instead of generating new insights. Worse: it creates a false sense of customer proximity — “We did an Empathy Map after all” [6].

Solution: Conduct at least 5 qualitative interviews before the workshop. If that’s not possible, use existing data sources: support tickets, NPS comments, reviews, usability test recordings. Mark every Post-it that isn’t based on a concrete data source as a hypothesis.

2. Confusing the Empathy Map with a persona

Symptom: The Empathy Map is decorated with demographic data, hobbies, and a stock photo, then used as a “persona.”

Why it hurts: An Empathy Map is a situational snapshot — it describes how a user thinks, feels, and acts in a specific situation. A persona is a permanent reference profile with demographic and psychographic data, goals, frustrations, and a representative scenario [7]. Using the Empathy Map as a persona creates a superficial profile that is neither situational enough for design decisions nor permanent enough as a reference document.

Solution: Use the Empathy Map as an intermediate step: Research → Empathy Map → Persona. The Empathy Map synthesizes research data; the persona consolidates multiple Empathy Maps into a permanent profile.

3. Formulating emotions too generically

Symptom: The Feels quadrant reads “frustrated,” “satisfied,” “uncertain” — without context.

Why it hurts: Generic emotion labels create no empathy and provide no design impulses. “Frustrated” can mean a hundred different things. What exactly frustrates the user? At what moment? Triggered by what?

Solution: Formulate emotions as sentences with context: “Anxious because no confirmation arrives for 3 days after filing the claim” instead of “uncertain.” Link every emotion to a specific trigger.

4. Only considering the frontstage user

Symptom: The Empathy Map is created only for the end customer — but not for the claims processor, the field representative, or the partner delivering the service.

Why it hurts: In service design, frontstage and backstage are inseparably connected. If the claims processor works with a system that doesn’t show the complete claims history, the customer experiences having to explain their problem repeatedly — even though the processor isn’t at fault. Only empathizing with the customer overlooks the backstage causes of frontstage problems.

Solution: For complex services, create at least two Empathy Maps: one for the end customer and one for the most important backstage actor. Multi-stakeholder design offers further protocols for working with multiple perspectives.

5. Creating the map and then forgetting it

Symptom: The Empathy Map is created in a workshop, photographed, uploaded to a Confluence board — and never looked at again. No follow-up, no transfer into journey maps or personas.

Why it hurts: The Empathy Map has no intrinsic value. Its value comes from the decisions it informs. Without transferring into subsequent tools (personas, journey maps, service blueprints), it remains a colorful poster.

Solution: Plan the transfer as an explicit next step. At the end of the workshop, define: “The top 3 insights will be transferred into the journey map by [date], responsible: [name].”

When the Empathy Map does NOT work

1. For complex multi-stakeholder services: If your service has 5+ different stakeholder groups (e.g., hospital: patients, doctors, nurses, administration, IT, health insurers), a single Empathy Map per stakeholder is too shallow. Use the combination of Stakeholder Map + Journey Map or a multi-stakeholder design protocol instead.

2. When quantitative data should replace the qualitative phase: The Empathy Map is a qualitative tool. If you have 10,000 support tickets, text mining is more efficient than Post-its. Use the Empathy Map as a complement — to enrich quantitative patterns with qualitative depth.

3. For radical innovation: If the user or the context doesn’t exist yet (e.g., a completely new service concept for a new market), you can’t create an Empathy Map based on existing research. Speculative methods (Design Fiction, Future-State Journey Mapping) are better suited here.

4. As a substitute for real user research: The Empathy Map synthesizes research — it doesn’t replace it. A 60-minute workshop with the internal team generates no customer knowledge if no user research has taken place beforehand.

Frequently asked questions

What is an Empathy Map?

An Empathy Map is a visual tool that divides a user’s perspective into structured segments: what the person says, thinks, does, and feels, plus their pains (obstacles/frustrations) and gains (wishes/benefits). It was developed by Dave Gray at XPLANE and popularized through Osterwalder’s Business Model Generation (2010). In service design, it serves to synthesize research data and align teams on a shared customer perspective.

How do you create an Empathy Map?

In four phases: (1) Define user and context — specific, not generic. (2) Review research data — interviews, support tickets, observations. (3) Fill segments — 5 minutes per segment (Says, Thinks, Does, Feels, Pains, Gains). (4) Identify patterns and formulate top 3 insights. Plan for 60 minutes for a complete workshop. Important: Without prior user research, the map is based on assumptions, not insights.

What is the difference between an Empathy Map and a persona?

An Empathy Map is a situational snapshot — it describes how a user thinks, feels, and acts in a specific situation. A persona is a permanent reference profile with demographic and psychographic data, goals, and frustrations. The Empathy Map is a workshop artifact (lifespan: hours to days), the persona is a reference document (lifespan: months to years). Ideally, create Empathy Maps first and then consolidate them into personas.

Which version of the Empathy Map should I use?

There are two versions: the original (2010) with 6 segments (Think & Feel, See, Hear, Say & Do, Pains, Gains) and the 2017 revision with 4 quadrants (Says, Thinks, Does, Feels) plus Goal. In practice, a hybrid is recommended: the four quadrants plus Pains and Gains as separate fields. The original works best for broad exploration, the revision for specific situational analysis.

Can I create an Empathy Map without interviews?

Technically yes, but you shouldn’t. An Empathy Map without a data foundation reflects the team’s assumptions, not the customer’s reality. If interviews aren’t possible, use existing data sources: support tickets, reviews, NPS comments, usability tests, social media comments. Mark every Post-it not based on a concrete data source as a hypothesis.

How many Empathy Maps do I need?

One per user group and context. If you have two customer segments (e.g., individual and business customers), create two maps. If you want to cover frontstage and backstage, additionally create a map for the most important internal actor (e.g., the claims processor). More than 4–5 maps become unwieldy — prioritize the most important segments.

Related methods

A typical sequence in service design: With user research you collect the data foundation. With the Empathy Map you synthesize the data into a user perspective. With personas you consolidate permanent reference profiles. With the Customer Journey Map you map the chronological experience. With the Service Blueprint you connect frontstage and backstage.

• User Research in Service Design: When you don’t have a data foundation yet — user research provides the inputs an Empathy Map needs
• Personas and Prototypes: When you want to turn the situational Empathy Map into a permanent reference profile
• Customer Journey Mapping: When you want to move from a snapshot to a chronological process map
• Service Blueprint: When you want to make the backstage processes behind the customer experience visible
• Service Design Methods: Overview: When you want to understand how the Empathy Map fits into the overall picture of service design tools

---

Research methodology

This article synthesizes findings from Dave Gray’s original format and his 2017 revision, Osterwalder’s integration into Business Model Generation (2010), Kouprie and Visser’s academic framework on empathy in design (2009), Gasparini’s empirical investigation of the Empathy Map as a learning tool (2015), and the analysis of 8 German-language expert contributions on the Empathy Map.

Limitations: Academic literature on the Empathy Map is limited — it is primarily a practitioner tool without formal validation studies. Most sources are practice guides, not empirical studies. The practical example (insurance claims process) is illustratively constructed, not a documented case study.

Disclosure

SI Labs offers consulting services in the area of service innovation. In the exploration phase of the Integrated Service Development Process (iSEP), we use the Empathy Map to structure customer perspectives before transferring them into journey maps and service blueprints. This practical experience informs the framing of the method in this article. Readers should be aware of potential perspective bias.

References

[1] Gray, Dave. “Empathy Map Canvas.” XPLANE / Gamestorming, 2010. https://gamestorming.com/empathy-mapping/ [Practitioner | Original Tool Creator | Quality: 85/100]

[2] Osterwalder, Alexander, and Yves Pigneur. Business Model Generation. Hoboken, NJ: Wiley, 2010. [Book | Popularization of Empathy Map | Citations: 15,000+ | Quality: 90/100]

[3] Gray, Dave. “Updated Empathy Map Canvas.” Medium, 2017. https://medium.com/the-xplane-collection/updated-empathy-map-canvas-46df22df3c8a [Practitioner | Updated Version by Original Creator | Quality: 80/100]

[4] Bland, David J., and Alexander Osterwalder. Testing Business Ideas. Hoboken, NJ: Wiley, 2020. [Book | Hybrid Format Integration | Citations: 500+ | Quality: 82/100]

[5] Kouprie, Merlijn, and Froukje Sleeswijk Visser. “A framework for empathy in design: stepping into and out of the user’s life.” Journal of Engineering Design 20, no. 5 (2009): 437-448. DOI: 10.1080/09544820902875033 [Journal Article | Theoretical Foundation | Citations: 800+ | Quality: 85/100]

[6] Gasparini, Isabela. “Empathy Map for Software Engineering Students: A Case Study.” International Journal of Engineering Education 31, no. 6 (2015): 1735-1744. [Journal Article | Empirical Study | Quality: 72/100]

[7] Cooper, Alan. The Inmates Are Running the Asylum: Why High-Tech Products Drive Us Crazy and How to Restore the Sanity. Indianapolis: Sams Publishing, 1999. [Book | Persona Method Origin | Citations: 5,000+ | Quality: 90/100]