Activity ⚡ High

Mental Model Mapping

Draw how users think before you design what they see — map the model in their head, not the model in yours.

User Research Information Architecture Navigation Design Concept Validation Discovery Research Cognitive Research

01 — TL;DR

Two sentences.

Mental Model Mapping is a qualitative user research activity — most fully developed by Indi Young in her book Mental Models: Aligning Design Strategy with Human Behavior (2008) — in which users are interviewed about their goals, behaviours, and thinking processes in a specific domain (not about a specific product), and their responses are documented and organised into a visual map showing the tasks, thoughts, and feelings that make up their mental model of that domain. The design team then aligns the product's current capabilities below that map to identify gaps, mismatches, and opportunities. It is the most direct method for answering the foundational design question that determines whether a product feels intuitive or confusing: does the structure, vocabulary, and logic of the product match the structure, vocabulary, and logic that users already have in their heads before they encounter it?

The method's defining distinction from persona research or standard user interviews is its domain focus rather than its product focus. Mental Model Mapping asks users about their goals and behaviours in a domain (managing their finances, planning a trip, learning a new skill) rather than about their experience with a specific product — which reveals the mental model users bring to any product in that domain, including the one being designed. That domain focus is what makes the resulting maps generative rather than merely evaluative: a mental model map reveals not just what is wrong with the current product but what a product in that domain would need to look like to feel genuinely intuitive to the users whose mental models have been mapped.

02 — When to Use

Apply this when…

→ An information architecture or navigation redesign is being planned and the team needs to understand how users conceptualise the content domain before designing the structure

→ A new product is being designed in an established domain (personal finance, health tracking, project management) and the team needs to understand the mental model users already have of that domain

→ Users are consistently using vocabulary different from the product's labels — the vocabulary mismatch reveals a mental model mismatch the method can characterise precisely

→ Multiple user segments may have fundamentally different mental models of the same domain — different models imply different product structures

→ A product is generating systematic user confusion that usability testing has not fully explained — the cause may be a conceptual model mismatch rather than an interface problem

→ A competitor's product is being evaluated and the team wants to understand what mental model it has established in shared users

When NOT to apply it

Skip it when the research question is about a specific product's usability — Mental Model Mapping is a domain research method, not a product evaluation method. Skip it when the team is time-constrained and needs quick directional research — full Mental Model Mapping requires significant time investment, and Card Sorting or Tree Testing provide faster IA-relevant input. And skip it when the domain is too narrow or too well-defined for meaningful conceptual variation — the method is most valuable in broad domains where users' conceptual structures vary meaningfully.

03 — How It Works

The mechanism

Mental Model Mapping works by documenting what is in users' heads about a domain — the tasks they perform, the thoughts they have while performing them, the feelings those tasks produce, and the tools and resources they use — and organising this documentation into a visual structure that reveals patterns across users. The visual map makes explicit the conceptual structure that users are implicitly applying when they engage with any product in the domain, providing design teams with a precise target for alignment.

Indi Young's mental model methodology

Indi Young's methodology, developed through consulting work with major technology companies in the early 2000s and documented in Mental Models (2008), adapts the academic mental model research tradition into a practical design research method. Young's approach begins with behavioural interviews — conversations about what users do and think in a domain, not about what they want from a product — and processes the interview data through a systematic coding and visualisation process. Interview responses are coded as tasks (things the user does), considerations (things the user thinks about), and guiding principles (values and beliefs). The coded data is organised into a tall, segmented diagram showing mental space towers arranged horizontally, with the product's current capabilities aligned below to show coverage gaps.

The interview approach and the alignment map

Mental Model Mapping runs in two phases. The research phase conducts behavioural interviews with 8–15 users in the relevant domain — conversations that ask users to describe what they do, think, and feel when engaging with a goal or task in the domain, explicitly avoiding product-specific questions. The synthesis phase processes interview transcripts to extract task statements, codes them into mental space towers, and creates the visual alignment map. The alignment map places users' mental space towers across the top and the product's current features and content across the bottom — making gaps (towers with no product support) and misalignments (product features with no corresponding mental space) immediately visible as design opportunities.

The method deliberately excludes product experience from the interviews

The most counterintuitive Mental Model Mapping principle is the deliberate exclusion of product-specific questions from the research interviews. Most UX research asks users about their experience with a specific product; Mental Model Mapping asks only about their goals, behaviours, and thinking in the relevant domain without reference to any product. This exclusion is the method's defining characteristic. By documenting the domain mental model without product contamination, the researcher captures what users bring to any product before it shapes their thinking — which provides the design team with an uncontaminated view of the conceptual structure they need to match. If the interviews asked about product experience, the resulting map would document users' model of the product, not their model of the domain.

Mental space coverage and model alignment score

Mental Model Mapping produces two primary analytical outputs. Mental space coverage measures what proportion of users' mental space towers are addressed by the product's current capabilities — gaps in coverage represent design opportunities where the product could serve users better by addressing currently unsupported aspects of their domain mental model. Model alignment score evaluates whether the product's structural logic (its navigation, its terminology, its conceptual model) matches the structural logic of users' mental model — high alignment means users can transfer their domain knowledge to the product with minimal learning; low alignment means the product requires users to build a new model before they can use it effectively.

Mental Model Mapping vs. Card Sorting — complementary, not interchangeable

Both address users' conceptual structures, but they answer different questions. Card Sorting asks users to organise a specific set of content items — revealing how users group and categorise the items being sorted. Mental Model Mapping asks users to describe their goals and behaviours in a domain — revealing the full conceptual structure of their domain understanding, including tasks, considerations, and feelings that Card Sorting's predefined item set cannot capture. Card Sorting is a faster, more targeted method for testing specific IA hypotheses; Mental Model Mapping is a more comprehensive, more time-intensive method for discovering the full domain mental model from scratch. Use them in sequence: Mental Model Mapping to discover the model, Card Sorting to test specific organisational hypotheses derived from it.

04 — Real Example

Rosenfeld Media's mental model map that restructured a newspaper website

Lou Rosenfeld's consulting work with a major online newspaper provides one of the most clearly documented Mental Model Mapping applications in information architecture practice. The newspaper's editorial team had organised their website around the newspaper's internal section structure — World News, National News, Business, Sports, Entertainment — which reflected the newspaper's production workflow rather than how readers thought about finding news.

Mental Model Mapping interviews with readers revealed a fundamentally different organisational logic: readers thought about news in terms of active stories they were following ("what is happening with the trade negotiations?"), topic areas of personal interest (technology, climate, local politics), and time relevance ("what happened today? what happened this week?"). These three dimensions — story, topic, and time — were not represented in the editorial section structure, which organised by production category rather than by reader intent. The mental model map showing the gap between readers' domain model and the product's current organisation provided direct evidence for a significant navigation redesign. The resulting redesign — which introduced story-following features, topic-based browsing, and time-filtered views alongside the existing section structure — produced measurable improvements in session depth and return visits by directly addressing the mental model gaps the research had identified.

Rosenfeld Media · Newspaper Website

Mental Model Map revealed a three-dimensional reader model that production-category navigation could not address

Aligning IA to readers' domain model lifted session depth and return visits

05 — In the Wild

Test yourself & see real examples

No examples yet — be the first.

Spotted a product whose navigation and structure clearly matches how you think about the domain — where everything is where you expected it to be — or one whose structure makes you feel like you need to learn a new way of thinking about something you already understand? Submit what you observed. Every approved example gets attributed to you.

✓ Reviewed before publishing ✓ Your name on every example you submit ✓ Violation or fix — both welcome

Seen a product that matches your domain mental model — or one that seems organised for the company's convenience rather than the user's thinking? Help grow the evidence base.

06 — Common Mistakes

Where teams go wrong

✗ Asking about product experience rather than domain behaviour. The most common mistake is contaminating the interviews with product-specific questions. "How do you use our product to manage your finances?" captures the user's model of the product, not their model of the domain. The interview must ask exclusively about domain behaviour: "Tell me about the last time you reviewed your monthly spending. What did you look at? What were you thinking about? What were you trying to understand?" That domain-behaviour framing is what produces the uncontaminated mental model that makes the map genuinely useful as a design target.

✗ Mental space towers that are too broad or too narrow. Towers labelled "managing money" or "staying healthy" are too broad to be actionable. Towers labelled "checking the joint savings balance on a Tuesday morning" are too narrow to reveal patterns. The right granularity captures a coherent cluster of related tasks performed with related considerations — specific enough to imply specific design responses, broad enough to represent a pattern across multiple users rather than one user's idiosyncratic behaviour.

✗ Using the alignment map as a final deliverable rather than a living design tool. The map's value comes from using it as an ongoing reference throughout design — not as a deck slide produced and archived. It should be physically present in design reviews, used to evaluate whether proposed IA decisions increase or decrease alignment, whether new features are addressing mental space gaps or adding to already well-covered areas, and whether the product's vocabulary is moving toward or away from the words that appear in the towers.

✗ Conducting too few interviews to reveal meaningful patterns. A map built on four or five interviews may reflect individual idiosyncrasies rather than shared domain mental models — accurate for those participants but not generalisable. The method typically requires 8–15 interviews for a well-bounded domain, more for diverse user segments. Too-few-interviews maps feel authoritative — full visual structure, complete-looking towers — but lack the evidentiary base to justify treating them as design direction.

07 — Variations & Related Principles

Connected ideas

Mental Model Mapping is the comprehensive cognitive research method for understanding users' domain conceptual structures. Its closest relationships are with the research methods that address related but narrower questions.

The most important pairing is Mental Model Mapping with information architecture design. The method's output — a visual map of users' domain mental model — is the most direct possible input to IA decisions, because IA is fundamentally about organising information in structures that match users' conceptual models of the content domain. Teams that design IA without a mental model map are choosing between intuition (their own mental model) and convention (structures used by competitors) — neither of which may match users' actual domain models. A mental model map replaces those indirect inputs with direct evidence of the conceptual structure the IA needs to support.

08 — 10-Min Exercise

Run it right now

⏱ 10 minutes · Solo · No prep

The Domain Interview Sketch

1. Pick a domain your product operates in — not your product itself, but the broader domain. If you make a project management tool, the domain is "managing work and teams." If you make a personal finance app, the domain is "managing personal money."

2. Write five behavioural interview questions about this domain that do not mention your product or any specific product. Each question should ask about what users do, think, or feel:

— "Tell me about the last time you [domain task]. What did you do first?"

— "What do you find most stressful about [domain activity]?"

— "When you are [domain goal], what information do you find yourself looking for?"

— "How do you currently [domain task] — walk me through what you actually do?"

— "What does [domain concept] mean to you? How would you explain it to someone?"

3. Write the questions without any reference to your product. If you find it difficult to ask about the domain without mentioning your product, that is a signal — your team's thinking about the domain may be so product-contaminated that you are finding it hard to imagine users' relationship with the domain independently of your product.

4. Identify one person who represents your target user and plan to ask them these five questions. Their answers are the raw material for a mental model map — and even a single informal conversation using domain-behaviour questions will reveal aspects of users' domain thinking that product-focused questions miss.

10 minutes