Concept Map: Definition, Step-by-Step Guide & Examples for Service Design

A concept map is a visual representation that shows concepts as nodes and the relationships between them as labeled connecting lines. Unlike simple brainstorming or mind mapping, a concept map forces you to explicitly name the type of relationship between two concepts — “causes,” “requires,” “enables” — making hidden assumptions visible [1].

The method was developed in 1972 by Joseph D. Novak at Cornell University to track changes in children’s conceptual understanding. Its theoretical foundation is David Ausubel’s theory of meaningful learning: new knowledge is permanently anchored when it is consciously linked to existing knowledge [1][2]. What began as a learning tool has long since evolved into a planning and analysis instrument for organizations — from knowledge architecture at NASA to information architecture at Sun Microsystems [1][3].

Search for “concept map” online and you will find two types of results: tool vendors (Miro, Lucidchart, Canva) trying to sell their software, and educational portals explaining classroom exercises. None of the results show how to use a concept map in a service design project — for instance, to untangle the web of relationships between customer segments, touchpoints, and backend systems at an insurance company. None explain the difference between a concept map and a mind map clearly enough that you can choose the right method in your next project meeting. And none honestly name the situations where a concept map is the wrong choice.

This guide fills those gaps.

From Cornell to CmapTools: Where the Method Comes From

Joseph D. Novak (born 1932) was a professor of education and biology at Cornell University. In 1972, he led a twelve-year research program tracking how children acquire scientific concepts. The problem: interview transcripts were so extensive that the research team could not reliably identify changes in children’s understanding. Novak developed the concept map as a solution — a visual format that made a learner’s conceptual structure mappable and comparable [1].

The theoretical basis came from David Ausubel’s theory of meaningful learning, published in 1963. Ausubel’s central thesis: new knowledge is only permanently integrated when the learner consciously links it to existing knowledge — as opposed to rote learning, which is quickly forgotten and cannot be transferred to new problems [2]. The concept map makes this linking process visible.

From 1987 onward, Novak collaborated with Alberto J. Canas at the Florida Institute for Human and Machine Cognition (IHMC). Together they developed CmapTools — free software that made concept mapping accessible for research, education, and organizations [4]. CmapTools is still used worldwide today, from schools to NASA, which employs the tool for knowledge management in space missions [1].

What Distinguishes a Concept Map from Other Visualizations

Martin J. Eppler compared four visualization formats in a widely cited 2006 study: concept maps, mind maps, conceptual diagrams, and visual metaphors. His finding: each format has a specific area of strength, and combining multiple formats produces the greatest benefit for knowledge construction and sharing [5].

Format	Structure	Strength	Weakness
Concept Map	Network with labeled connections	Making relationships explicit, connecting knowledge domains	Time-consuming, requires practice
Mind Map	Tree structure with central node	Quick brainstorming, idea collection	No relationship labels, flat structure
Service Blueprint	Process flow with swimlanes	Temporal sequences and interfaces	No conceptual relationships
Stakeholder Map	Circle model with proximity/influence	Making actors and their power visible	No content relationships between actors

The crucial distinction: A mind map organizes ideas around a topic — radially, associatively, without relationship labels. A concept map connects concepts to each other — as a network, with labeled connections, and allows cross-links between different branches. If you want to know which ideas belong to a topic, use a mind map. If you want to understand how those ideas relate, use a concept map [5].

When to Use a Concept Map

A concept map is particularly valuable when you need to understand, communicate, or analyze a web of relationships between concepts — not when you want to collect ideas or document a process flow.

Use a concept map when:

• You need to structure a complex topic where many terms exist in non-linear relationships — such as the dependencies between customer segments, products, channels, and backend systems at an insurance company
• A cross-functional team has different mental models of the same system and you need to create shared understanding — the labeled connections force clarity where discussions remain vague
• You need to externalize expert knowledge before a knowledge holder leaves the company or a project is handed over [1]
• You are in the Discover phase of a service design project, mapping the problem space before diving deep with Customer Journey Mapping or Empathy Maps
• You want to identify cross-links between knowledge domains — according to Novak, these cross-links are the most creative part of the concept map [1]

Use a different tool when:

Situation	Better Alternative	Why
You want to collect ideas quickly without defining relationships	Brainstorming / Mind Map	Mind maps are faster and require no relationship labels
You need a temporal process flow with touchpoints	Customer Journey Map	Journey maps capture the chronological sequence of the customer experience
You want to document front- and backstage processes of a service	Service Blueprint	Blueprints reveal the invisible infrastructure behind the customer interface
You need a root cause analysis for a specific problem	Ishikawa Diagram	Ishikawa structures causes into categories, not concept relationships
You want to position stakeholders by influence and interest	Stakeholder Map	Stakeholder maps show power dynamics, not concept relationships

How to Create a Concept Map: Step-by-Step Guide

Time frame: 90 to 120 minutes for a first version with a team. Expect 2 to 3 revision rounds over several days until the map stabilizes [1][3].

Materials: Whiteboard or large paper surface, sticky notes (one color for concepts, another for linking words), markers. Digital: CmapTools (free), Miro, or Mural.

Step 1: Formulate a Focus Question (10 Minutes)

Every concept map answers a focus question. Without this question, the map becomes an aimless collection of terms [1].

Poorly formulated	Well formulated
”What do we know about onboarding?"	"What concepts and dependencies determine the new-user experience in our digital onboarding?"
"Our product portfolio"	"How do our products, customer segments, and distribution channels relate to each other?"
"Service design"	"What methods and artifacts connect user research to the implementation of new services?”

Practical tip: The focus question should begin with “What,” “How,” or “Why.” If you cannot formulate the question in one sentence, your scope is too broad — split the topic into two separate maps.

Step 2: Identify Key Concepts (15 Minutes)

List 15 to 25 concepts relevant to the focus question. Write each concept on a separate sticky note [1].

Sources for concepts:

• Interviews with subject-matter experts and stakeholders
• Existing documentation (project reports, process descriptions)
• Results from User Research
• Glossaries and terminology lists
• Team brainstorming

Sort concepts hierarchically: Arrange them from most general to most specific. General, overarching concepts will go at the top of the map, specific ones at the bottom [1].

Place the sorted sticky notes as a “parking lot” next to your work surface. You will not use all of them — and you will add new ones during the process.

Step 3: Build the Initial Structure (20 Minutes)

Place the most general concept at the top of the work surface. Arrange the next-most-specific concepts below it. Connect related concepts with lines.

Critical — the labeled connections: Every connecting line needs one or more words describing the relationship. Two concepts plus the linking words form a proposition — a meaningful statement [1].

Examples of propositions:

• “Customer segment” — defines requirements for — “Touchpoint”
• “Onboarding process” — consists of — “Identity verification”
• “Complaint management” — influences — “Customer satisfaction”

Common mistake: Connecting lines without labels. An unlabeled line says: “These two concepts are somehow related.” That is worthless. Force yourself to use a verb or preposition. If you cannot label the connection, the relationship is either unclear or nonexistent [1].

Step 4: Identify Cross-Links (15 Minutes)

Cross-links connect concepts from different areas of the map. They are the most valuable part of a concept map because they reveal connections that are lost in linear representations [1].

Deliberately search for connections between concepts that are far apart in the hierarchy or in different “branches” of the map. Typical questions:

• “Is there a direct dependency between this concept on the left and that one on the right?”
• “If this concept changes — which other concepts are affected?”
• “Which connections surprise us?”

Visually distinguish cross-links (e.g., dashed line, different color) so they are recognizable at first glance.

Step 5: Revise and Refine (20 to 30 Minutes)

A good concept map does not emerge on the first attempt. Novak and Canas recommend at least three revision rounds [1]. During the redesign work for java.sun.com at Sun Microsystems, the team went through 53 numbered versions before the final map was complete [3].

Check in each round:

• Do the propositions hold up? Read each connection as a sentence: “Concept A — [linking word] — Concept B.” Does the sentence make sense?
• Are concepts missing? Are there aspects of the focus question not yet covered?
• Are concepts redundant? Does the same concept appear under two different names?
• Are there additional cross-links? Draw those you missed in Step 4.

Step 6: Validate and Share (10 to 15 Minutes)

Present the concept map to a subject-matter expert or stakeholder who was not involved in the creation process. Their feedback reveals blind spots the creation team cannot see [3].

Documentation: Photograph the physical map or export the digital version as a PDF. Add a brief companion document with:

• The focus question
• The list of participants
• Key insights (especially surprising cross-links)
• Open questions that emerged from the map

Example: Concept Map for Insurance Onboarding

Context: A product manager at an insurance company needs to improve digital onboarding for new customers after contract signing. Several departments — IT, customer service, compliance, sales — have different ideas about what “good onboarding” means. Before the customer journey is worked out in detail, the team creates a concept map to clarify the conceptual field of “onboarding.”

Focus question: “What concepts and dependencies determine the new-user experience after contract signing?”

Identified concepts (selection): New customer, Contract signing, Welcome package, App registration, Identity verification (KYC), Primary contact person, Self-service portal, FAQ section, Product configuration, First claim, Customer satisfaction, Churn risk, Compliance requirements, Data privacy, Sales promises

Selected propositions:

Concept A	Linking words	Concept B
New customer	goes through	App registration
App registration	requires	Identity verification (KYC)
Identity verification	is subject to	Compliance requirements
Compliance requirements	can delay	New-user experience
Sales promises	creates expectations for	New-user experience
New-user experience	influences	Customer satisfaction
Customer satisfaction	correlates negatively with	Churn risk
Self-service portal	complements	Primary contact person
Welcome package	refers to	Self-service portal
Welcome package	refers to	Primary contact person

Cross-links (surprising connections):

• “Sales promises” — is in tension with — “Compliance requirements”: Sales promises fast onboarding, but KYC verification takes an average of 48 hours. This expectation gap was previously invisible to any team.
• “FAQ section” — does not cover — “First claim”: New customers who file a claim within the first week find no help in the FAQ — because the FAQ is designed for existing customers.
• “Data privacy” — restricts — “Primary contact person”: The primary contact person cannot access contract data without additional authentication — making the first interaction cumbersome.

Result: The team identified three previously invisible dependencies that no one had noticed in their departmental silos. The concept map became the foundation for the subsequent journey mapping work and helped structure collaboration between sales, compliance, and IT.

Note: This example is illustratively constructed to demonstrate the method in an insurance context.

Variations and Advanced Techniques

Hierarchical Concept Map (Novak Standard Format)

The classic form: general concepts at the top, specific ones at the bottom, cross-links across. Best suited for analyzing knowledge domains and understanding hierarchies [1]. Use this variation when your topic has a natural hierarchy — such as organizational structures, product categories, or regulatory requirements.

Spider Map

A central concept in the middle, all others radiating outward. Similar to a mind map but with labeled connections. Suitable for exploring a single concept with many facets — such as “customer satisfaction” with all its influencing factors.

Flowchart Concept Map

Combines concept map elements with a process flow. Concepts are arranged in chronological order, relationships show both logical and temporal dependencies. Suitable for representing service processes where you want to show both the sequence and the conceptual relationships.

Systemic Concept Map

Extends the classic map with feedback loops and interactions. Suitable for complex systems where cause and effect are not linear — such as ecosystems, market dynamics, or organizational change. Compare the Cynefin Framework, which helps assess a system’s complexity before choosing the appropriate visualization method.

Collaborative Concept Mapping in Team Settings

For use in cross-functional teams, a facilitated workshop is recommended:

1. Everyone creates an individual mini-map (5 to 8 concepts, 10 minutes)
2. Compare maps — where do mental models align, where do they diverge?
3. Discuss differences — each divergence is a potential insight
4. Create a consensus map — document the shared understanding as a reference

This method makes implicit knowledge visible and uncovers misunderstandings before they become project problems.

Concept Map Template: Quick-Start Checklist

Preparation

• Focus question formulated as a concrete question (not a topic)
• 15 to 25 key concepts identified
• Concepts sorted hierarchically (general to specific)
• Materials prepared (sticky notes, markers, or digital tool)
• Participants invited (3 to 6 people from different departments)

Creation

• Most general concept placed at top
• More specific concepts arranged below
• All connections labeled with words (no unlabeled arrows)
• Each proposition readable as a sentence: “Concept A — [verb] — Concept B”
• Cross-links between different areas identified
• Cross-links visually highlighted (different color or dashed)

Quality Check

• At least one revision round completed
• Propositions checked for factual accuracy
• Focus question is answered by the map
• Map validated by someone not involved in creation
• Companion document created (focus question, participants, insights, open questions)

When a Concept Map Does NOT Work

Concept maps are not the right tool for every visualization problem. Know these limitations:

1. Simple, linear relationships: If your topic has a clear temporal or logical sequence (A then B then C), a flowchart or customer journey map is more efficient. The network structure of a concept map adds unnecessary complexity to linear topics.

2. Rapid idea collection: If you want to collect as many ideas as possible in 15 minutes, a concept map is too slow. Labeling every connection takes time. Use brainstorming or a mind map instead, and transfer the results into a concept map later when relationships become relevant.

3. Large teams without facilitation: Beyond 7 to 8 people, collaborative concept mapping without experienced facilitation becomes chaotic. Discussions about linking words can become endless. In large groups, use the individual-map method (everyone first creates their own map) or split the team into smaller groups.

4. When the audience does not know the method: A concept map is only a communication tool if the recipient can read it. If you are presenting results to stakeholders who have never seen a concept map, translate the insights into a more familiar format — a table or a presentation with the key propositions as bullet points.

5. Quantitative data: Concept maps are suited for qualitative relationships between concepts, not for data visualization. If you want to display numbers, frequencies, or correlations, use charts, dashboards, or a value stream map.

Research confirms these limitations: Ruiz-Primo and Shavelson (1996) documented that concept maps are less reliable with learners lacking prior experience, and that quality depends heavily on the expertise of the creators [6].

5 Common Mistakes in Concept Mapping

1. No Focus Question Defined

Symptom: The map grows in all directions, nobody knows when it is “done,” and the discussion jumps between unrelated topics.

Solution: Formulate the focus question before you write down the first concept. Every concept you add must contribute to answering this question. If it does not, it belongs in a separate map.

2. Connections Without Labels

Symptom: Lines between concepts without words. This is not a concept map — it is a relationship diagram without informational value.

Solution: Every connection needs at least one word. If you cannot find an appropriate label, the relationship is either unclear (clarify it) or nonexistent (remove the line) [1].

3. Too Many Concepts at Once

Symptom: 50+ concepts on one map, illegible text, overlapping lines. The map communicates nothing.

Solution: Limit the map to 15 to 25 concepts. For complex topics: create an overview map with 10 to 15 main concepts and develop individual areas in separate detail maps (sub-maps). This is what NASA did for their space mission projects [1].

4. Mind Map Format with Concept Map Aspirations

Symptom: All concepts radiate from a center, there are no cross-links, no hierarchy, no labeled connections.

Solution: If you are creating a tree structure, consciously use a mind map (that is a good tool). If you want to create a concept map, you need: hierarchical structure, labeled connections, and cross-links. The difference is methodologically grounded, not aesthetic [5].

5. Created Once, Never Revised

Symptom: The map hangs on the wall and is never touched. New insights do not flow in.

Solution: Plan revision rounds as fixed appointments. At Sun Microsystems, the concept mapping process for java.sun.com took 49 weeks — with 53 numbered versions, 36 individual interviews, and feedback from 10 external reviewers [3]. Your project does not need comparable effort, but plan for at least two to three iterations.

Concept Maps in Service Design Practice

In practice, concept maps prove most valuable at three points in service design projects:

1. Discover phase: Mapping the problem space. Before diving deep with user research, a concept map helps you understand the web of relationships between actors, systems, and processes. It defines the scope of your research and identifies where the largest knowledge gaps lie.

2. Define phase: Consolidating knowledge. After interviews and observations, you have many individual findings. A concept map helps assemble these findings into a coherent picture — especially when different user groups (see service design user types) provide different perspectives.

3. Handover phases: Ensuring knowledge transfer. When a project team changes or results are handed to an implementation team, implicit knowledge is lost. A concept map documents not just what the team knows, but how the knowledge elements connect — that is more valuable for handovers than a presentation with bullet points.

Concept maps do not replace service blueprints, journey maps, or empathy maps. They complement them — as a meta-layer showing how a project’s individual artifacts relate to each other. Cavalieri et al. (2021) demonstrated this approach in a research project on product-service system design: a concept map served as a planning instrument to ensure that no relevant information about PSS elements was overlooked in early design phases [7].

Frequently Asked Questions

What is a concept map?

A concept map is a visual representation of knowledge that shows concepts as nodes and the relationships between them as labeled connecting lines. Two concepts plus the linking words form a proposition — a readable statement such as “Customer satisfaction — influences — Churn risk.” The method was developed in 1972 by Joseph D. Novak at Cornell University.

What is the difference between a concept map and a mind map?

Three main differences: (1) Structure — a mind map radiates tree-like from a center, a concept map is a network with cross-links. (2) Labels — mind map branches have no relationship labels, concept map connections are always labeled. (3) Purpose — mind maps collect associations around a topic, concept maps analyze relationships between concepts. For quick brainstorming, use a mind map. For deep understanding, use a concept map.

How do I create a concept map?

In six steps: (1) Formulate a focus question. (2) Identify 15 to 25 key concepts and sort them hierarchically. (3) Build the initial structure — general concepts at top, specific ones below, label all connections. (4) Identify cross-links between different areas. (5) Complete at least two to three revision rounds. (6) Have someone not involved in creation validate the map. Allow 90 to 120 minutes for the first version.

What tools are suitable for concept maps?

For physical workshops: whiteboard and sticky notes (fastest for group work). For digital creation: CmapTools (free, specifically designed for concept maps), Miro and Mural (broader functionality, good for remote teams), Lucidchart (professional diagrams). For individual work, a sheet of paper and a pen will also suffice.

When should I use a concept map instead of a mind map?

Use a concept map when you want to understand the type of relationship between concepts — not just which concepts belong together. Typical use cases: mapping knowledge domains, creating cross-functional alignment, externalizing expert knowledge, understanding problem spaces in service design projects. Use a mind map when you want to quickly collect ideas or exploratively open up a topic.

How many concepts should a concept map have?

For starters, Novak and Canas recommend 15 to 25 concepts. Fewer than 10 concepts represent the topic too superficially. More than 30 concepts make the map unreadable. For complex topics: create an overview map with 10 to 15 main concepts and develop individual areas in separate detail maps.

Related Methods

A typical sequence in a service design project: Use the concept map to map the problem space and build shared understanding. Deepen knowledge about users through user research. Document the customer experience with the customer journey map. Plan implementation with the service blueprint.

• Stakeholder Map: When you want to represent actor relationships (influence, interest, dependencies) rather than concept relationships
• Empathy Map: When you want to understand a single person’s perspective (thinks, feels, says, does) rather than the overall system
• Design Thinking: For the overall process in which the concept map is embedded as a tool in the Discover and Define phases
• Ishikawa Diagram: When you want to analyze the causes of a specific problem rather than map a conceptual field
• Brainstorming: When you want to quickly collect ideas before concept mapping, then transfer them into a map

---

Research Methodology

This article synthesizes findings from the foundational work of Novak and Canas on concept mapping (1972-2008), Eppler’s comparative study on visualization formats (2006), Nesbit and Adesope’s meta-analysis on learning effectiveness (2006, N=55 studies, 5,818 participants), the Dubberly Design Office case study on concept map application at Sun Microsystems, and Cavalieri’s PSS research (2021). Sources were selected based on methodological rigor, practical relevance, and recency.

Limitations: The academic literature on concept maps predominantly comes from educational contexts. Empirical studies on systematic application in service development are limited. The practice example (insurance onboarding) is illustratively constructed, not a documented case study. The effectiveness data from Nesbit and Adesope’s meta-analysis relates to learning contexts, not organizational applications.

Disclosure

SI Labs offers consulting services in the area of service innovation and uses concept maps as a tool in the Discover phase of the Integrated Service Development Process (iSEP). This practical experience informs the positioning of the method in this article. Readers should be aware of the potential perspective bias.

Bibliography

[1] Novak, Joseph D. and Alberto J. Canas. “The Theory Underlying Concept Maps and How to Construct and Use Them.” Technical Report IHMC CmapTools 2006-01 Rev 01-2008, Florida Institute for Human and Machine Cognition, 2008. https://cmap.ihmc.us/docs/theory-of-concept-maps [Technical Report | Foundational Work | Citations: 10,000+ | Quality: 95/100]

[2] Ausubel, David P. The Psychology of Meaningful Verbal Learning. New York: Grune & Stratton, 1963. [Foundational Work | Learning Theory | Citations: 15,000+ | Quality: 90/100]

[3] Dubberly, Hugh and Paul Pangaro. “Using Concept Maps in Product Development: Preparing to Redesign java.sun.com.” Dubberly Design Office, 2009. https://www.dubberly.com/articles/using-concept-maps-in-product-development.html [Case Study | Product Development | Quality: 80/100]

[4] Novak, Joseph D. and Alberto J. Canas. “The Origins of the Concept Mapping Tool and the Continuing Evolution of the Tool.” Information Visualization 5 (2006): 175-184. DOI: 10.1057/palgrave.ivs.9500126 [Journal Article | Citations: 500+ | Quality: 85/100]

[5] Eppler, Martin J. “A Comparison between Concept Maps, Mind Maps, Conceptual Diagrams, and Visual Metaphors as Complementary Tools for Knowledge Construction and Sharing.” Information Visualization 5 (2006): 202-210. DOI: 10.1057/palgrave.ivs.9500131 [Journal Article | Comparative Study | Citations: 800+ | Quality: 85/100]

[6] Ruiz-Primo, Maria Araceli and Richard J. Shavelson. “Problems and Issues in the Use of Concept Maps in Science Assessment.” Journal of Research in Science Teaching 33, no. 6 (1996): 569-600. DOI: 10.1002/(SICI)1098-2736(199608)33:6<569::AID-TEA1>3.0.CO;2-M [Journal Article | Citations: 800+ | Quality: 80/100]

[7] Cavalieri, Lorenzo et al. “A concept map to support the planning and evaluation of artifacts in the initial phases of PSS design.” Research in Engineering Design 32 (2021): 177-199. DOI: 10.1007/s00163-021-00358-9 [Journal Article | PSS Design | Quality: 78/100]