{"site":{"name":"Koji","description":"AI-native customer research platform that helps teams conduct, analyze, and synthesize customer interviews at scale.","url":"https://www.koji.so","contentTypes":["blog","documentation"],"lastUpdated":"2026-05-25T15:57:45.224Z"},"content":[{"type":"documentation","id":"1dac87ad-8849-4c10-9c2c-cb624fd25ea6","slug":"empathy-map-guide","title":"Empathy Map: The Complete Guide to Building User Empathy","url":"https://www.koji.so/docs/empathy-map-guide","summary":"An empathy map is a 6-section collaborative visualization tool that organizes user research data into Says, Thinks, Does, Feels, Pains, and Gains sections. Created by Dave Gray at XPLANE and published in Gamestorming (2010), it helps product teams build shared understanding of users. AI platforms like Koji auto-generate empathy map data from interviews by thematically tagging quotes and detecting sentiment across all participants.","content":"# Empathy Map: The Complete Guide to Building User Empathy\n\nAn empathy map is one of the most powerful and underutilized tools in UX research. It takes raw user data — interviews, observations, survey responses — and organizes it into a shared visual framework that helps entire teams understand the humans they are designing for. Done well, it is the difference between a product team that makes decisions based on assumptions and one that makes decisions grounded in genuine user understanding.\n\nAs Sarah Gibbons of Nielsen Norman Group puts it: **\"Empathy is important because the best possible experiences are designed with human users in mind.\"** The empathy map is the mechanism that makes that principle concrete.\n\nThis guide covers everything you need to know about empathy maps: their origin, the full 6-section framework, a step-by-step creation process, common mistakes, and how AI-powered interview tools like Koji help you build richer empathy maps faster.\n\n---\n\n## What Is an Empathy Map?\n\nAn empathy map is a collaborative visualization tool used to articulate what a team knows about a particular type of user. It organizes observations about users into a structured canvas with sections covering what they say, think, do, and feel — along with their pains and gains.\n\nThe empathy map serves two purposes:\n\n1. **Individual understanding** — Helping a single researcher or designer synthesize what they learned from user research\n2. **Shared understanding** — Aligning a cross-functional team around a consistent, human-centered view of the user\n\nResearch shows that **75% of customers are more loyal to brands that understand them on a deeper level**, and 73% cite customer experience as a key factor in brand loyalty decisions. The empathy map is the foundational tool for developing that understanding systematically.\n\n---\n\n## The Origin of the Empathy Map\n\nThe empathy map was created in the mid-2000s by the design consultancy **XPLANE**. The tool originated from work by Scott Matthews, XPLANE's Creative Director, who began drawing caricature-style sketches — called \"big heads\" — to capture information about the people they were designing for.\n\nDave Gray, founder and chairman of XPLANE, refined Matthews' concept into the structured empathy map format. As Gray explains, the tool was designed to help teams develop **\"deep, shared understanding and empathy for other people\"** — and teams soon discovered uses far beyond product design: improving customer experience, navigating organizational change, designing better work environments.\n\nThe Empathy Map was formally published in 2010 through **\"Gamestorming,\"** the influential facilitation playbook co-authored by Sunni Brown and James Macanufo. This gave the tool wide visibility in the design, agile, and innovation communities. It was subsequently adopted by the Stanford d.school, Accenture, Nielsen Norman Group, and IDEO as a core design thinking tool.\n\nIn 2017, Dave Gray released an updated version of the empathy map that integrated **Pains** and **Gains** directly into the canvas, transforming it from a 4-quadrant to a 6-section framework more suitable for modern product and service design.\n\n---\n\n## The 6-Section Empathy Map Framework\n\nThe modern empathy map consists of 4 quadrants arranged around a central user figure, plus 2 additional sections at the bottom. Each section captures a distinct dimension of the user's experience.\n\n### The User (Center)\n\nAt the center of the map is the user or persona being studied. Include:\n- A name or persona label (e.g., \"Marketing Manager Maya\")\n- The situation or context being explored (e.g., \"Evaluating a new analytics tool\")\n- Their primary goal in this context\n\n### Quadrant 1: SAYS\n\nWhat the user says out loud — verbatim quotes from interviews, usability sessions, or survey open-ends.\n\n- Direct quotes only (no paraphrasing)\n- Captures articulated needs, concerns, and reactions\n- The richest source: user interviews with verbatim transcripts\n\n**Examples:**\n- \"I wish this process did not take so long.\"\n- \"I need to know what the next step is before I can commit.\"\n- \"I keep getting lost after the third screen.\"\n\n### Quadrant 2: THINKS\n\nWhat the user is thinking throughout their experience — including what they might not say out loud.\n\n- Inferred from behavioral cues, body language, pauses, and follow-up probing\n- Reveals internal concerns, mental models, and unstated assumptions\n- Sometimes contradicts what users say (Says ≠ Thinks is a valuable signal)\n\n**Examples:**\n- Thinking: \"Will I break something if I click this?\" (even if they say nothing)\n- Thinking: \"Am I the only one who finds this confusing?\"\n- Thinking: \"This looks like it will take too long to set up.\"\n\n### Quadrant 3: DOES\n\nWhat actions the user takes — their observable behaviors, interaction patterns, and workarounds.\n\n- Sourced from usability session recordings, analytics data, session replays\n- Captures actual behavior, which often diverges from stated behavior\n- Reveals coping strategies and workarounds users have invented\n\n**Examples:**\n- Returns to the home page whenever confused about where they are\n- Opens a second browser tab to compare options side by side\n- Screenshots the confirmation screen instead of downloading the PDF\n\n### Quadrant 4: FEELS\n\nThe emotional state and mood of the user throughout their experience.\n\n- Captured through tone of voice, word choice, facial expressions, and body language\n- Note both the emotion and its trigger (e.g., \"frustrated when asked to re-enter data\")\n- Positive and negative emotions are equally important\n\n**Examples:**\n- Anxious when entering payment information\n- Relieved when the progress bar shows how far along they are\n- Confused by the terminology in error messages\n- Delighted when the system remembers their preferences\n\n### Section 5: PAINS (Bottom Left)\n\nThe blockers, fears, frustrations, and obstacles that prevent the user from achieving their goals.\n\n- What creates friction or prevents task completion?\n- What are they afraid of getting wrong?\n- What obstacles do they encounter repeatedly?\n\n**Examples:**\n- Fear of committing to a subscription before understanding the value\n- Confusion about which tier of service applies to their use case\n- Time pressure from a manager expecting results the same day\n\n### Section 6: GAINS (Bottom Right)\n\nThe outcomes, aspirations, and results the user wants to achieve — what success looks like for them.\n\n- What are they trying to accomplish?\n- What would make their day meaningfully better?\n- What would they tell their manager they achieved?\n\n**Examples:**\n- Wants to produce a report in under 30 minutes\n- Wants to feel confident recommending a tool to their team\n- Wants to avoid looking incompetent in front of stakeholders\n\n---\n\n## How to Create an Empathy Map: Step-by-Step\n\n### Step 1: Define Your Scope\n\nDecide whether you are creating a one-user empathy map (based on one interview or observation session) or an aggregated empathy map (synthesizing patterns across multiple users). Both are valid — the aggregated version is more common and more useful for product decisions.\n\n### Step 2: Gather Your Research Data\n\nCollect the raw material you will synthesize into the map:\n- Interview transcripts and recordings\n- Session notes and observations\n- Survey open-ended responses\n- Analytics data and session recordings\n- Support tickets and sales call notes\n\nThe quality of your empathy map is directly proportional to the quality of your underlying research. Rich, exploratory user interviews — not just surveys — are the best source of Says, Thinks, and Feels data.\n\n### Step 3: Set Up Your Map\n\nDraw or use a digital template with the 6 sections. Popular tools include Miro, FigJam, or a physical whiteboard. Label the sections: Says, Thinks, Does, Feels, Pains, Gains.\n\n### Step 4: Populate With Sticky Notes\n\nWith your research data in hand (transcripts, notes, recordings), write each observation on a separate sticky note and place it in the appropriate section.\n\n**Key principles:**\n- Keep each sticky to one observation\n- Be specific — generic observations are not useful\n- Use the user's own words when possible in the Says section\n- Add the source (e.g., \"Interview P3, 12 min\") for traceability\n\n### Step 5: Synthesize and Cluster\n\nStep back and look for patterns:\n- Which observations appear in multiple quadrants? (e.g., user says \"it seems complicated\" AND shows confused body language AND clicks the wrong button)\n- Where do SAYS and THINKS contradict each other?\n- Which PAINS are most frequent or severe?\n- Which GAINS are most strongly desired?\n\nCluster related sticky notes and label each cluster with a theme.\n\n### Step 6: Identify Insights and Opportunities\n\nThe empathy map itself is not the output — the insights derived from it are. Ask:\n- What does this tell us about unmet needs?\n- Where is the greatest pain that we could reduce?\n- What gain could we unlock that users most desire?\n- Where does the experience most fail to match what users expect?\n\nDocument these insights and connect them to your product roadmap or design backlog.\n\n---\n\n## How AI-Powered Interviews Give You Better Empathy Map Data\n\nThe richest empathy map data comes from user interviews — specifically the Says, Thinks, and Feels quadrants. The challenge: traditional interview-based research is slow, expensive, and hard to scale.\n\nTraditional tools like moderated interview platforms require scheduling, a trained moderator, manual transcription, and hours of synthesis. Teams that cannot afford to invest that time often skip the qualitative research entirely — and build products based on assumptions.\n\nKoji solves this with AI-moderated interviews that conduct themselves. Here is how Koji enriches each section of the empathy map:\n\n| Empathy Map Section | Traditional Approach | With Koji AI |\n|---------------------|---------------------|-------------|\n| **SAYS** | Manual transcription of recorded interviews | Auto-transcribed verbatim quotes, tagged by theme |\n| **THINKS** | Inferred by trained moderator during live session | AI probing questions uncover unstated concerns |\n| **DOES** | Behavioral observation in usability sessions | Structured yes/no and ranking questions reveal behavioral patterns |\n| **FEELS** | Tone analysis by experienced researcher | Sentiment analysis across all interviews, with emotional language flagged |\n| **PAINS** | Synthesized manually from themes | Automatic theme detection across 10s or 100s of interviews |\n| **GAINS** | Surfaced through ranking and priority questions | Ranking question type surfaces explicit desired outcomes |\n\nKoji's 6 structured question types — open-ended, scale, single choice, multiple choice, ranking, and yes/no — map directly onto empathy map needs:\n- **Open-ended questions** capture Says and Thinks\n- **Scale questions** quantify emotional intensity (Feels)\n- **Yes/no and single choice** document behavioral patterns (Does)\n- **Ranking questions** surface prioritized Gains\n\nResearch shows that **prioritizing collaboration increases product quality by 34% and customer satisfaction by 41%** — and empathy maps are the shared artifact that enables that collaboration. With Koji, your entire team can review the empathy map data in a shared research dashboard the day after fielding your study, not two weeks later.\n\n---\n\n## Common Empathy Map Mistakes to Avoid\n\n### 1. Mixing observations and interpretations\n\nThe Says quadrant should contain verbatim quotes, not your interpretation of what users meant. Keep raw observations and interpretations separate.\n\n### 2. Building from assumptions, not data\n\nAn empathy map built from guesses rather than actual user research is worse than useless — it creates false confidence. Every entry should be traceable to a specific research source.\n\n### 3. Creating it in isolation\n\nEmpathy maps are most valuable as collaborative artifacts. Build them in workshops with product, design, and engineering together — not as a solo research deliverable.\n\n### 4. Treating the map as the endpoint\n\nThe empathy map is a means to an end. The goal is the insights and design opportunities that emerge from it. If you build the map but do not draw conclusions, you have done the hard part without getting the value.\n\n### 5. Never updating it\n\nUser needs evolve. An empathy map built from research conducted 18 months ago may be outdated. Build a practice of refreshing empathy maps when launching new features or entering new market segments.\n\n---\n\n## Empathy Maps vs. User Personas\n\nEmpathy maps and personas are complementary, not interchangeable:\n\n| | Empathy Map | User Persona |\n|---|-------------|-------------|\n| **Focus** | Emotional and behavioral depth | Representative user profile |\n| **Input** | Qualitative observations from research | Synthesized patterns across user segments |\n| **Output** | Shared understanding of a specific experience | Reusable reference character |\n| **Best used for** | Synthesizing research sessions | Communicating user segments to stakeholders |\n| **Relationship** | Feeds into persona creation | Informed by empathy maps |\n\nThe typical workflow: conduct user interviews → build empathy maps from each interview → synthesize across maps → create personas.\n\n---\n\n## Related Resources\n\n- [User Interview Guide: How to Conduct Effective Research Interviews](/docs/user-interview-guide)\n- [Thematic Analysis: How to Find Patterns in Qualitative Data](/docs/thematic-analysis-guide)\n- [Generative Research: How to Uncover User Needs You Did Not Know Existed](/docs/generative-research-guide)\n- [Structured Questions Guide: Mixing Qualitative and Quantitative Research](/docs/structured-questions-guide)\n- [How to Analyze User Interview Data](/docs/how-to-analyze-user-interview-data)\n- [User Research Report Template](/docs/ux-research-report-template)\n\n## Further reading on the blog\n\n- [Best AI Market Research Tools in 2026: The Complete Buyer's Guide](/blog/ai-market-research-tools-2026) — AI has fundamentally changed market research. This guide compares the leading AI market research platforms—from AI-native interview tools li\n- [B2B Customer Research: The Complete Guide for Product Teams (2026)](/blog/b2b-customer-research-guide-2026) — B2B customer research is harder than B2C — you are navigating buying groups of 10+ stakeholders, gatekeepers, and enterprise procurement cyc\n- [Best AI Customer Interview Tools in 2026: The Complete Buyer's Guide](/blog/best-ai-customer-interview-tools-2026) — AI has fundamentally changed how product teams conduct customer research. Here are the best AI customer interview tools in 2026 — ranked by \n\n<!-- further-reading:blog -->\n","category":"Research Methods","lastModified":"2026-05-13T00:26:36.807295+00:00","metaTitle":"Empathy Map: Complete Guide to Building User Empathy (2026)","metaDescription":"Learn how to create an empathy map with the complete 6-section framework (Says, Thinks, Does, Feels, Pains, Gains). Includes step-by-step process and how Koji AI interviews give you richer empathy data.","keywords":["empathy map","empathy mapping","empathy map ux","how to create an empathy map","empathy map template","empathy mapping user research","design thinking empathy map"],"aiSummary":"An empathy map is a 6-section collaborative visualization tool that organizes user research data into Says, Thinks, Does, Feels, Pains, and Gains sections. Created by Dave Gray at XPLANE and published in Gamestorming (2010), it helps product teams build shared understanding of users. AI platforms like Koji auto-generate empathy map data from interviews by thematically tagging quotes and detecting sentiment across all participants."}],"pagination":{"total":1,"returned":1,"offset":0}}