Design thinking

Design (Fashion / Product / Communication / UI-UX) (B.Des) · Foundations · free study notes for exams, placements and interviews

In short

Design thinking is a structured, human-centred way of solving ill-defined ("wicked") problems: you understand real people first, frame the right problem, generate many options, build rough prototypes and test them with users before committing resources. It is not a style of drawing but a *process of reasoning* — abductive, iterative and evidence-driven — and it applies equally to a garment range, a water purifier, a poster campaign or a payments app. In B.Des Foundation it is examined as the Design Process paper (stages, tools, case studies, short projects) at NID, NIFT, MIT-ID, Srishti, Pearl and university B.Des programmes, and in placement rounds at Titan, Godrej, Tata Elxsi, Zoho, Swiggy, Zomato, Flipkart, Infosys Design, Wipro Digital, TCS Interactive and design studios. Interviewers almost never ask "define design thinking" alone — they ask you to walk your portfolio project through its stages and defend why each decision was taken.

Explained

What it is and where it came from

Definition: a solution-seeking methodology that begins with human needs, treats the problem statement itself as a variable, and uses cycles of making and testing to converge on a desirable, feasible and viable answer.
Wicked problems (Horst Rittel, 1970s): problems with no single correct formulation, no stopping rule and no right/wrong answer — only better or worse. Design thinking exists for these; engineering optimisation handles well-defined problems.
Lineage: Herbert Simon, *The Sciences of the Artificial* (1969) — design as a science of making; Rittel and Webber on wicked problems; Bruce Archer and the Design Methods movement; Nigel Cross on designerly ways of knowing; Rolf Faste and David Kelley teaching it at Stanford; IDEO commercialising it; Tim Brown's *Change by Design* (2009) and Roger Martin's *The Design of Business* (abductive logic, "knowing vs. making").
Three logics: *deductive* (general → specific, proof), *inductive* (observations → pattern), abductive (observed outcome → the most plausible explanation or "what might be"). Design thinking is primarily abductive — it proposes what does not yet exist.
Three lenses that must overlap (IDEO): desirability (do people want it?), feasibility (can it be made with available technology and skill?), viability (can it sustain itself financially?). A solution strong on only one or two lenses fails.
Human-centred design (HCD) is the parent philosophy; design thinking is its operational process. ISO 9241-210 (Ergonomics of human-system interaction — human-centred design for interactive systems, adopted in India as IS/ISO 9241-210) codifies the same loop: understand context of use → specify user requirements → produce design solutions → evaluate against requirements.

The five-stage model (Stanford d.school / IDEO) — most examined

Empathise — understand the user in their own context. Methods: non-participant and participant observation, shadowing, contextual enquiry, semi-structured interviews (open questions, "tell me about the last time…", laddering with "why?" five times), diary studies, cultural probes, immersion (NID's classic village/craft immersion), extreme-user interviews, secondary research. Output: raw field notes, photographs, verbatim quotes, artefact inventories.
Define — frame the problem. Synthesis tools: affinity mapping (cluster sticky notes into themes), empathy map (says / thinks / does / feels), persona (a composite, evidence-based user — not an invented friend), user journey map with touchpoints and pain points, stakeholder map, POEMS framework (People, Objects, Environments, Messages, Services). Output: a point-of-view (POV) statement — *[user] needs [need] because [surprising insight]* — plus How-Might-We (HMW) questions pitched at the right altitude (not so broad they give no traction, not so narrow they contain the answer).
Ideate — generate many options before judging any. Brainstorming rules (Osborn's four — defer judgement, go for quantity, encourage wild ideas, build on others' ideas — extended by IDEO's seven, which add one conversation at a time, be visual, stay on topic). Tools: mind mapping, SCAMPER (Substitute, Combine, Adapt, Modify, Put to another use, Eliminate, Reverse), morphological chart, brainwriting / 6-3-5 (6 people × 3 ideas × 5 rounds = 90 ideas; the classic 6 rounds gives 108), worst possible idea, analogous and biomimetic inspiration, Crazy 8s (8 sketches in 8 minutes), bodystorming. Then converge by dot voting, a weighted Pugh matrix or now/wow/how grid.
Prototype — make ideas tangible cheaply. Fidelity ladder: sketch → paper wireframe → storyboard → foam/clay/mock-up → Wizard of Oz (human fakes the system) → clickable prototype (Figma) → working MVP. Each prototype should answer one specific question ("will users find the filter?"), not demonstrate the whole product.
Test — learn, not prove. Usability testing with 5–8 users per round (most severe issues surface by the fifth user), think-aloud protocol, task-success rate, time on task, error count, System Usability Scale (SUS) questionnaire, A/B testing at scale, Wizard-of-Oz trials, pilot deployment. Feed findings back to any earlier stage — the loop is non-linear.

Other process models you must be able to compare

Double Diamond (UK Design Council, 2005; refreshed 2019): Discover → Define → Develop → Deliver. Diamond 1 finds the *right problem*, diamond 2 finds the *right solution*; each diamond opens (divergent thinking) then closes (convergent thinking). The 2019 version adds design principles, methods bank, leadership and engagement.
IDEO's three spaces: Inspiration → Ideation → Implementation — overlapping, not sequential.
Google Design Sprint (Jake Knapp): a time-boxed five-day version — Map, Sketch, Decide, Prototype, Test — used widely in Indian product teams.
Lean Startup / Lean UX: Build → Measure → Learn, validated learning, MVP, pivot. Complements design thinking on the viability side.
Agile / Scrum: delivery framework (sprints, backlog, stand-ups); design thinking supplies the *what and why*, Agile the *how fast*.
Engineering design process (AICTE "Design Thinking and Innovation" course, common in year 1–2): need identification → specification → concept generation → embodiment → detail design → manufacture, with design-for-X (manufacture, assembly, environment) and a strong emphasis on requirement specifications.
Jugaad innovation (Radjou, Prabhu, Ahuja) and Gandhian / frugal engineering: improvise, do more with less, think and act flexibly, include the margin, follow your heart. The constraint-led Indian cousin of design thinking — exam favourite for case questions.

Model	Stages	Strength	Watch out
d.school 5-stage	Empathise, Define, Ideate, Prototype, Test	Easy to teach and examine	Looks linear if taught badly
Double Diamond	Discover, Define, Develop, Deliver	Shows diverge/converge clearly	Less explicit about empathy tools
IDEO 3 spaces	Inspiration, Ideation, Implementation	Honest about overlap	Vague for a student report
Design Sprint	Map, Sketch, Decide, Prototype, Test	Fast, decision-forcing	Thin user research
Lean Startup	Build, Measure, Learn	Market validation	Can skip framing the problem

Divergence, convergence and problem framing

Divergent thinking: generate breadth — many, varied, unexpected options; judgement suspended. Convergent thinking: evaluate, combine and select against criteria.
Never mix the two in the same session; a critic in an ideation room cuts idea counts sharply, which is why facilitation rules exist.
Problem framing is the highest-leverage act. Fixing the brief ("design a better lift") versus reframing it ("reduce the *perceived* wait") is the classic mirror-in-the-lift reframe.
Design fixation: early attachment to one concept (often the first sketch or a competitor's solution) suppresses alternatives. Antidotes: parallel prototyping, deliberately generating the worst idea, bringing in an outsider.
Brief vs. problem statement: a brief is given (client, constraints, deliverables, budget, timeline); the problem statement is what you *derive* from research and must be able to defend.

Research methods: qualitative and quantitative

Primary (you collect it): interviews, observation, surveys, co-design workshops, usability tests. Secondary: literature, market reports, census and NSS data, standards, competitor teardown.
Qualitative answers *why* — small n, rich depth, thematic analysis, no statistical claim. Quantitative answers *how many* — larger n, needs sampling and statistics.
Sampling: purposive/extreme-user sampling for qualitative work; random or stratified sampling for surveys. Never quote a percentage from 8 interviews.
Co-design / participatory design (Scandinavian tradition): users are design partners, not subjects — essential for craft-cluster, rural and accessibility projects.
Ethics and law in India: informed consent (written, in the participant's language), anonymisation, child and vulnerable-participant safeguards, institutional ethics clearance for human-subject studies. Personal data collected in user research falls under the Digital Personal Data Protection Act, 2023 — notice, purpose limitation, consent, data-principal rights; the DPDP Rules were notified in 2025 with phased commencement, so check the current compliance date before relying on a specific deadline. Accessibility obligations flow from the Rights of Persons with Disabilities Act, 2016 and its harmonised guidelines; government and public-service digital products follow the GIGW (Guidelines for Indian Government Websites and Apps) and WCAG-based accessibility levels.
Protecting outcomes: the Designs Act, 2000 protects the *aesthetic* configuration of an article (registration from the Design Office, Kolkata; 10 years, extendable by 5); functional inventions go under the Patents Act, 1970; brand marks under the Trade Marks Act, 1999; drawings, sketches and UI screens are protected as artistic works under the Copyright Act, 1957. A design dictated solely by function is not registrable as a design — a standard interview trap.

Measuring and defending a design decision

Criteria matrix / Pugh matrix: score concepts against weighted criteria relative to a datum; forces explicit trade-offs instead of taste.
Usability metrics: task success rate, time on task, error rate, SUS score (0–100; ≈68 is the long-standing average benchmark), Net Promoter Score, conversion rate, retention.
Cost of late change: fixing a problem gets dramatically more expensive as you move from sketch to tooling to market — the economic argument for cheap early prototypes.
Design ROI: make the claim with your own measured before/after numbers (drop-off, support tickets, returns), not with borrowed index figures whose methodology you cannot defend.
Limitations — know these for the "critique design thinking" question: can become a sticky-note ritual with no shipped outcome ("design theatre"); small samples over-generalised; weak on systems, policy, long time-horizons and second-order environmental consequences; assumes users can articulate latent needs; ignores power — whose problem gets framed; poorly suited to problems that are genuinely well-defined and need optimisation, not exploration. Counter-movements: systems thinking, speculative/critical design, transition design, decolonising design.

Key points

Design thinking - a human-centred, iterative process for ill-defined problems, not a drawing style.
Wicked problem - no single formulation, no stopping rule, no right answer; the reason the method exists.
Abduction - the designer's logic: reasoning to "what might be", unlike deduction and induction.
Desirability, feasibility, viability - the three overlapping lenses every solution must satisfy.
Five stages - Empathise, Define, Ideate, Prototype, Test; iterative, with loops back from any stage.
Double Diamond - Discover, Define, Develop, Deliver; diamond 1 = right problem, diamond 2 = right solution.
Divergent then convergent - widen options first, judge later; never in the same session.
POV statement - [user] needs [need] because [insight]; the output of Define.
HMW question - reframes the POV as an actionable opportunity at the right altitude.
Empathy map and journey map - synthesis tools turning raw field notes into pain points and opportunities.
Persona - an evidence-based composite user, built from research, never invented.
Brainstorming rules - defer judgement, quantity, wild ideas, build on others, one conversation, be visual.
SCAMPER / 6-3-5 / Crazy 8s - structured idea-generation tools that beat unstructured discussion.
Prototype fidelity - lowest fidelity that answers one specific question; rough is a feature, not a flaw.
Usability testing - 5–8 users per round with think-aloud; most severe issues surface early.
SUS - 10-item usability questionnaire scored 0–100; ≈68 is the commonly cited average.
Design fixation - premature attachment to one concept; countered by parallel prototyping.
Jugaad / frugal innovation - the Indian constraint-led variant: do more with less, include the margin.
Designs Act, 2000 - protects aesthetic configuration (10 + 5 years); function goes to patents, screens to copyright.
DPDP Act, 2023 - consent, notice and purpose limitation govern user-research data in India.

Formulas and examples

SUS score = 2.5 × [ Σ(odd items − 1) + Σ(5 − even items) ]
Odd items 1,3,5,7,9 are positively worded; even items 2,4,6,8,10 negatively worded; each raw response 1–5 (Strongly disagree → Strongly agree). Result is 0–100 and is a *percentile-like score, not a percentage*.

Problems found in usability testing = 1 − (1 − p)ⁿ
p = probability one user reveals a given problem (≈0.31 in Nielsen–Landauer's data), n = number of users tested. Dimensionless fraction.

6-3-5 brainwriting yield = P × I × R
P = participants, I = ideas written per round, R = rounds. Unit: ideas.

Weighted concept score S = Σ (wᵢ × rᵢ)
wᵢ = weight of criterion i (weights sum to 1), rᵢ = rating of the concept on criterion i (say 1–5). Unit: dimensionless score.

Interview sample effort E = n × (t_session + t_travel + t_notes)
n = participants, t in hours → E in person-hours; add transcription and synthesis time separately.

Example 1 — SUS score of a kirana billing app. Ten users; mean responses: Q1 4, Q2 2, Q3 4, Q4 2, Q5 4, Q6 2, Q7 5, Q8 1, Q9 4, Q10 3.

Odd items minus 1: (4−1)+(4−1)+(4−1)+(5−1)+(4−1) = 3+3+3+4+3 = 16.
Even items, 5 minus score: (5−2)+(5−2)+(5−2)+(5−1)+(5−3) = 3+3+3+4+2 = 15.
Sum = 31 → SUS = 2.5 × 31 = 77.5. Above the ≈68 benchmark, so usability is better than average — but with n = 10 the confidence interval is wide, so report it as indicative.

Example 2 — how many users to test. With p = 0.31, fraction of problems found:

n = 5 → 1 − 0.69⁵ = 1 − 0.1565 = 0.843 (84.3%)
n = 8 → 1 − 0.69⁸ = 1 − 0.0514 = 0.949 (94.9%)
n = 15 → 1 − 0.69¹⁵ = 1 − 0.0038 = 0.996 (99.6%)

Going from 5 to 15 users triples the cost to gain about 15 percentage points — so run three rounds of 5 with a redesign between rounds instead of one round of 15.

Example 3 — idea yield and cost of an ideation workshop. A studio runs 6-3-5 with 6 designers, 3 ideas each, 5 rounds → 6 × 3 × 5 = 90 ideas in about 30 minutes. Dot voting shortlists 8; a weighted matrix picks 3 for prototyping. Cost at a billing rate of ₹1,200 per designer-hour for 0.5 hour: 6 × 0.5 × 1,200 = ₹3,600, i.e. ₹40 per idea generated.

Example 4 — weighted concept selection for a water-bottle redesign. Criteria and weights: user need 0.35, manufacturability 0.25, cost 0.20, sustainability 0.20.

Concept	User need (0.35)	Manufacturability (0.25)	Cost (0.20)	Sustainability (0.20)	Weighted score
A — double-wall steel	5	3	2	4	3.70
B — single-wall steel	4	5	4	4	4.25
C — recycled PET	3	4	5	3	3.65

A: 0.35×5 + 0.25×3 + 0.20×2 + 0.20×4 = 1.75 + 0.75 + 0.40 + 0.80 = 3.70
B: 1.40 + 1.25 + 0.80 + 0.80 = 4.25
C: 1.05 + 1.00 + 1.00 + 0.60 = 3.65

Concept B wins. Sensitivity check: raise user need to 0.50 and drop manufacturability to 0.10 → A = 2.50+0.30+0.40+0.80 = 4.00, B = 2.00+0.50+0.80+0.80 = 4.10; B still leads, so the choice is robust.

Example 5 — research effort and A/B test read-out. Fieldwork: 12 contextual interviews in Pune, 1.5 h session + 1 h travel + 0.5 h notes = 3 h each → 12 × 3 = 36 person-hours, plus 12 h synthesis = 48 h ≈ 6 working days for one researcher. Participant honorarium ₹500 each = ₹6,000. Resulting checkout redesign tested A/B: control 4,000 sessions, 240 orders = 6.0%; variant 4,000 sessions, 300 orders = 7.5% — a relative lift of (7.5−6.0)/6.0 = 25%. Added orders at ₹900 average order value = 60 × 900 = ₹54,000 in the test window; report it with the test duration and a significance check, not as a permanent uplift.

Where it is used

Product and industrial design: Titan, Godrej Appliances, Tata Elxsi, Whirlpool India, Bajaj, TVS, Cello — concept clinics, user trials, ergonomics for Indian anthropometry, DFM reviews with vendors.
UI/UX and product teams: Zoho, Flipkart, Swiggy, Zomato, Razorpay, Paytm, PhonePe, CRED, Infosys Design, Wipro Digital, TCS Interactive — personas, journey maps, Figma prototypes, usability labs, A/B tests; UPI and DigiLocker flows are standard Indian HCD case studies.
Fashion and lifestyle: NIFT-style range building — consumer segmentation, mood and trend boards, muslin toiles as prototypes, fit sessions as user tests; Fabindia, Raymond, Aditya Birla Fashion, Myntra, craft-cluster co-design with artisans.
Communication design and social impact: behaviour-change campaigns for Swachh Bharat, immunisation, road safety; signage and wayfinding for metros and railways; message testing with low-literacy audiences.
Public services and rural innovation: Smart City dashboards, Aadhaar-enabled service counters, low-cost medical devices (Jaipur Foot, affordable ECG and neonatal warmers), Mitticool fridge, Atal Innovation Mission and Tinkering Labs, IIT/NID incubators.
Job roles: UX researcher, interaction designer, service designer, design strategist, product designer, innovation consultant (Deloitte, Accenture Song, EY), CMF designer, design manager.

Interview questions

Q: Is design thinking linear? A: No — the stages are a teaching sequence; in practice a test result can send you back to ideation or even to reframing the problem, and stages overlap.
Q: What exactly is the output of the Define stage? A: A defensible point-of-view statement — user, need and insight — plus How-Might-We questions; not a solution.
Q: How do you write a good HMW question? A: At the right altitude: broad enough to allow several solution directions, narrow enough to give traction, and free of any implied solution.
Q: Why prototype roughly instead of building the real thing? A: A rough prototype answers a specific question in hours at near-zero cost and invites criticism; a polished one invites approval and makes late changes expensive.
Q: How many users do you test with, and why? A: Five to eight per round with think-aloud — that reveals roughly 85–95% of severe issues — then redesign and test again, rather than one large round.
Q: What is design fixation and how do you avoid it? A: Locking onto an early concept so alternatives stop being explored; avoided by parallel prototyping, generating deliberately bad ideas and fresh reviewers.
Q: Difference between a persona and a user segment? A: A segment is a statistical market group (demographic or behavioural); a persona is a single research-based character capturing goals, context and pain points to design against.
Q: Is design thinking the same as Agile? A: No — design thinking decides what is worth building and why; Agile is a delivery framework for building and shipping it in increments. Teams run them together.
Q: What is abductive reasoning in design? A: Proposing the most plausible "what might be" from observed needs, rather than deducing from rules or inducing a pattern from data.
Q: Criticise design thinking. A: It can degenerate into workshop theatre with no shipped outcome, over-generalises from tiny samples, under-serves systemic and long-horizon problems, and ignores who gets to frame the problem.
Q: How do you handle a client who hands you a solution as the brief? A: Accept the brief, then run a short discovery to test the underlying need, and present evidence for either the brief or a reframe — with the business implication of each.
Q: Where does Jugaad fit? A: It is constraint-led frugal improvisation that reaches desirable outcomes with minimal resources — strong on feasibility and affordability, weak on repeatability, safety and scale unless formalised.
Q: What protects a design you create in India? A: The Designs Act, 2000 for aesthetic configuration, patents for functional inventions, copyright for drawings and screens, and trade marks for brand identity.

Common mistakes

Jumping to a solution in the Empathise stage and then using research only to justify it.
Quoting percentages from a handful of interviews, or calling 8 users a representative sample.
Inventing personas from imagination or stereotype instead of building them from field data.
Treating the stages as a one-way checklist to be ticked off, with no iteration and no evidence of looping back.
Over-polishing the first prototype, so critique becomes approval and changes become costly.
Asking leading questions in interviews ("wouldn't a bigger button help?") or asking users to predict future behaviour instead of describing past behaviour.
Writing a vague HMW question, or presenting a solution dressed up as a problem statement in a portfolio review.
Confusing a design registration with a patent, or assuming a copyright on a sketch protects the product's shape.
Submitting final renders with no process documentation — most B.Des juries and employers mark the process, not the gloss.