Design Thinking
Design Thinking: Five-phase human-centred innovation process β Empathise (understand users deeply), Define (frame the right problem), Ideate (generate many solutions), Prototype (build cheap representations), Test (learn from failure fast). The key insight: most innovation failures come from solving the wrong problem, not from poor execution of the right one.
What Is Design Thinking?β
Design Thinking emerged from Stanford's d.school (Hasso Plattner Institute of Design) and was popularised by the design firm IDEO, particularly through Tim Brown's 2009 book of the same name. It formalises the creative process that skilled designers use naturally into a repeatable, teachable methodology that non-designers can apply.
The central premise is that most innovation failures are problems of problem definition, not solution execution. Companies build products nobody wants, launch services that miss user needs, or implement organisational changes that create new problems while solving old ones β because they skipped or rushed the work of genuinely understanding the humans they're designing for.
Design Thinking is structured around three overlapping criteria for successful solutions: desirability (do people actually want this?), feasibility (can it be built?), and viability (does it sustain a business?). Most organisations default to feasibility and viability; Design Thinking insists that desirability must come first.
The five phases:
- Empathise: Immerse yourself in users' experiences β observe, interview, shadow. Build genuine understanding of their needs, motivations, and constraints beyond what surveys or analytics show.
- Define: Synthesise empathy research into a clear problem statement β the "Point of View" (POV) statement: [User] needs [need] because [surprising insight]. Frame the right problem before generating solutions.
- Ideate: Generate many diverse solutions β "How Might We?" questions, brainstorming, SCAMPER, reverse brainstorming. Quantity before quality; diverge before converging.
- Prototype: Build cheap, fast representations of ideas β paper mockups, role-play, Wizard of Oz prototypes. The goal is to learn, not to build; the prototype should be good enough to test, no better.
- Test: Put prototypes in front of real users. Observe their behaviour (not just their stated opinions). Fail fast; iterate. Return to earlier phases when the test reveals wrong assumptions.
How It Worksβ
Phase 1: EMPATHISE (days to weeks)
β Conduct user interviews (min. 5-7 per segment)
β Observe users in their natural environment
β "Empathy maps": what users say, think, do, feel
Phase 2: DEFINE (hours to days)
β Cluster insights from empathy research
β Identify patterns and tensions
β Write POV statement: "[User] needs [need] because [insight]"
β Test: is this the right problem to solve?
Phase 3: IDEATE (hours)
β "How Might We?" reframe of POV statement
β Diverge: generate 50-100+ ideas without evaluation
β Converge: cluster, vote, select 3-5 to prototype
Phase 4: PROTOTYPE (days)
β Build the cheapest possible representation
β Paper, foam, role-play, clickable mock, landing page
β Optimise for speed and learning, not quality
Phase 5: TEST (days)
β Test with users from the empathy phase
β Observe behaviour; ask "why?"
β Capture what works, what doesn't, what's surprising
β Return to Define or Ideate if needed
Three Real-World Examplesβ
IDEO and the Hospital Patient Experienceβ
IDEO was asked to redesign the patient experience in an Apria Healthcare hospital. Rather than immediately designing new spaces, the team spent days as "patients" β checking in with fake ailments, lying in beds, navigating corridors. Key empathy insight: the ceiling was the dominant visual experience for patients on gurneys or in beds; they spent hours staring at institutional tile and fluorescent lights.
This unexpected insight redefined the problem from "how do we make hospitals look better?" to "how do we create an experience patients spend hours staring at?" Solutions included redesigning ceiling panels, installing overhead art, and repositioning lighting. Without the empathy phase, this would never have been identified.
IBM's Design Thinking Transformationβ
In 2012, IBM committed to transforming 400+ development teams using Design Thinking. The key was the "Hills" framework β three declarative statements of who the user is, what they do, and what the outcome is β that replaced specification documents. Teams that adopted Design Thinking delivered products 30% faster with significantly higher NPS scores, according to IBM's internal studies. The speed increase came from the reduction of rework β building the right thing the first time instead of discovering misalignment in late-stage testing.
GE Healthcare MRI Scanner Redesignβ
GE Healthcare noticed that children undergoing MRI scans frequently needed sedation because the experience was terrifying β the machine was loud, enclosed, and industrial. Doug Dietz, a GE designer, applied Design Thinking. Empathy research with children revealed they were afraid of the machine and the medical context; the problem wasn't the MRI technology.
Reframe: "Create an adventure experience." Solution: themed MRI rooms β "pirate ship," "safari," "space adventure" β where the scan was narrated as part of the adventure. Patient experience scores improved dramatically; sedation rates for children dropped significantly. The same MRI hardware; completely different experience.
When to Use Itβ
β Design Thinking excels when:
- Solving problems where user needs are unclear or assumed
- Developing new products, services, or customer experiences
- Organisational challenges with a strong human component
- Any situation where previous "logical" solutions have failed to work in practice
β Less appropriate for:
- Well-defined technical problems with known solution spaces
- Situations requiring immediate decisions without research time
- Mature products where the problem is execution, not definition
| Pairs well with | Why |
|---|---|
| Jobs to Be Done | JTBD theory provides the analytical framework for the Empathise and Define phases |
| Minimum Viable Test | MVT is the Prototype + Test philosophy operationalised |
| SCAMPER | SCAMPER is a structured Ideate technique |
| First Principles | First Principles reconstructs solutions from user needs discovered in Empathise |
Common Misuses and Limitationsβ
Skipping the empathy phase. "We know our users" is the most dangerous assumption in innovation. Most failed products were built by people who were confident they understood user needs. The empathy phase must be genuine immersive research, not confirmation of existing beliefs.
Treating it as a linear process. Design Thinking is explicitly iterative β you will return to Define after a failed Test; you will jump between Ideate and Prototype rapidly. Teams that treat it as a checklist to complete in sequence miss the iterative heart of the method.
Prototype perfection. "We need to make the prototype look good before showing users" is a budget- and time-burning mistake. A Wizard-of-Oz prototype (a human manually doing what software will eventually do) or a paper sketch is often more informative than a polished mockup, because users give more honest feedback when the design doesn't look "done."
Confusing Design Thinking with design. Design Thinking is a problem-solving process applicable to any domain β healthcare, policy, organisational design, education. It is not exclusively about product aesthetics or UX. Its core contribution is human-centred problem definition, which is valuable everywhere.
Related Modelsβ
| Model | Relationship |
|---|---|
| Jobs to Be Done | Complementary frameworks; JTBD provides the theory of why users adopt solutions |
| Minimum Viable Test | MVT is Design Thinking's Prototype-Test philosophy in a product development context |
| Cynefin Framework | Design Thinking is appropriate for "complicated" and "complex" Cynefin domains |
| Reframing | The Define phase is a structured reframing of the initial problem |
Frequently Asked Questionsβ
How does Design Thinking differ from Agile development?
Design Thinking addresses what to build (problem definition and solution generation); Agile addresses how to build it (iterative execution). They're complementary: Design Thinking ideally precedes Agile development, defining validated user needs and solution directions that Agile teams then build and ship. Many teams that adopt Agile without Design Thinking end up iterating quickly toward the wrong goal. The combination β Design Thinking for problem definition, Agile for execution β is increasingly the standard approach in product development.
How many user interviews are needed in the Empathise phase?
Qualitative research guidelines suggest 5β7 participants per distinct user segment typically reveals ~80% of the key patterns. The goal is insight saturation β the point at which additional interviews stop generating new findings. More complex or diverse user populations may require 15β20 interviews. Nielsen Norman Group research supports 5 as the minimum for identifying major usability issues; Design Thinking generally requires more depth than usability testing, so 7β12 is a common target for the Empathise phase.
How do you know when to move from Prototype back to Define?
When users consistently misunderstand the prototype's purpose, the problem may be a misframed definition β you're solving the wrong problem. When users understand but don't care, you may have an accurate but unimportant problem. When users understand, care, but can't use the solution, you have an execution problem (stay in Prototype-Test). The test phase should explicitly check which of these failure modes is operating before deciding where to return in the process.
Further Readingβ
- Brown, T. (2009). Change by Design β IDEO's CEO on Design Thinking in business
- Kelley, D. & Kelley, T. (2013). Creative Confidence β the psychological foundations of Design Thinking
- Liedtka, J. & Ogilvie, T. (2011). Designing for Growth β Design Thinking applied to strategy
Apply with AIβ
π Run a Design Thinking session with MindMax β
This page is part of the MindMax Mental Models Knowledge Base.