Design is the disciplined transformation of an uncertain situation into a form people can use. It begins before the solution exists. Designers investigate the problem, understand the people and context, make constraints visible, generate alternatives, create representations, test assumptions and repeatedly alter the solution until the relationship between need, form and consequence becomes strong enough to release.
Design is therefore broader than styling. A beautiful object that is dangerous, inaccessible or impossible to maintain is not fully resolved design. A technically correct service that confuses users can also be badly designed.
The UK Design Council’s widely used Double Diamond describes design as alternating divergence and convergence through Discover, Define, Develop and Deliver. The Council explicitly notes that the model is not a rigid instruction manual and that real design iterates as teams learn. Source: Design Council.
Reading routes: begin with the child-friendly explanation; follow problem framing; explore research, idea generation, prototyping, testing, visual design, interaction, service design, accessibility, sustainability, ethics and the workshop.
Explain design to a child: do not decorate the wrong answer
Imagine a classroom has a box for returning library books, but younger children cannot lift the lid.
You could paint the box brighter. That changes appearance but not the problem.
You could lower the box, change the lid, create a slot or move it to a better location.
Design begins by asking what is actually stopping the user, then changes the system so the intended action becomes easier, safer or clearer.
1. Design begins before the brief is trusted
A client may ask for “an app.”
The real need might be reducing waiting time, clarifying information or allowing a task to happen outside office hours.
Starting with the requested artefact can lock the team into the wrong solution too early.
The first design move is often to question the brief.
2. Problems are framed, not simply found
The same situation can be framed differently.
“People keep missing appointments” can become a reminder problem, an accessibility problem, a scheduling problem, a transport problem or a trust problem.
Each frame generates different solutions.
Design therefore treats framing as a consequential decision.
3. A design problem contains a receiver
Who experiences the problem?
The answer can include users, operators, maintainers, regulators, bystanders and future owners.
A product convenient for the buyer can be difficult for staff to service.
Good design identifies the full receiver set before optimising one person’s experience.
4. Constraints create the design space
Budget, law, materials, dimensions, time, environment and safety restrict what can be built.
Constraints are not enemies of creativity.
They define the space in which creativity becomes useful.
A solution that ignores a hard constraint is an illustration, not a viable design.
5. Requirements and preferences should be separated
A wheelchair-accessible route may be a requirement.
A particular colour may be a preference.
Confusing the two causes teams to spend energy negotiating what cannot legitimately be traded away.
Design reviews should make constraint strength visible.
6. Discovery replaces assumption with observation
The first diamond in the Design Council model widens the problem through discovery before narrowing it into a definition. Source: Design Council Framework for Innovation.
Research can include interviews, observation, data, field visits, journey mapping and analysis of existing services.
The objective is not maximum research volume.
It is sufficient evidence to change the team’s mental model.
7. Interviews reveal experience but not perfect truth
People can explain frustrations, goals and context.
They can also misremember behaviour or describe what they wish they did.
Design research therefore combines reported experience with observation and other evidence where possible.
What users say and what users do are both data.
8. Observation exposes hidden work
A worker may tell a designer that a process takes three steps.
Observation reveals six handwritten checks, two unofficial spreadsheets and one phone call used to recover missing information.
These workarounds are often where the real design opportunity lives.
Design should study the current system before replacing it.
9. Journey maps reveal transitions
A service can appear good inside each department while feeling broken across the whole journey.
A journey map follows the user across stages: discover → decide → access → use → receive support → finish.
Handoffs reveal repeated questions, waiting and loss of state.
The map is useful because users experience the whole journey, not departmental boundaries.
10. Personas are representations, not fictional authorities
A persona can compress recurring user needs into a memorable reference.
It becomes dangerous when invented biography replaces evidence.
Strong personas are grounded in research and retain important variation.
Real users should not be forced to fit neat fictional categories.
11. Jobs-to-be-done thinking focuses on progress
People often “hire” a product or service to make progress in a situation.
A drill helps create a hole. A calendar helps coordinate time. A bank transfer helps discharge an obligation.
The useful design question is what progress the person is trying to make.
This can reveal alternatives outside the existing product category.
12. Defining the challenge is a convergence step
After discovery, teams need a bounded problem statement.
A useful definition identifies user, situation, need and constraints without embedding one predetermined solution.
“How might we reduce confusion during hospital discharge?” leaves more design space than “build a discharge app.”
Good definitions guide invention without prematurely closing it.
13. Divergence creates alternatives before commitment
Design improves when several plausible solutions are considered.
The first idea is often anchored to existing habits.
Divergence deliberately expands options.
The second diamond in the Double Diamond model widens again during development before convergence toward delivery. Source: Design Council.
14. Creativity needs comparison
An idea cannot be strong merely because it is novel.
It needs to be compared with alternatives against user value, feasibility, safety, cost and maintainability.
Design judgement is comparative.
The question is which option best fits the real constraint set.
15. Sketching is cheap thinking
A sketch externalises an idea before expensive commitment.
It allows a team to notice missing states, proportions or flows.
The sketch does not need aesthetic polish.
Its value lies in making assumptions discussable.
16. Storyboards design time, not only space
A product can look excellent in one static screen and fail across a sequence.
Storyboards show what happens before, during and after use.
They reveal transitions, emotion and information timing.
This is particularly useful for services and interactions.
17. Morphological thinking recombines solution dimensions
Original teaching model: imagine a delivery service with three options for handoff location, three notification methods and two identity-check methods.
The design space contains 3 × 3 × 2 = 18 basic combinations before considering other constraints.
Systematic combination can expose alternatives that brainstorming misses.
The point is not to build all 18. It is to understand the option space.
18. Prototypes are questions made tangible
A prototype should answer something.
A paper interface tests navigation. A foam model tests size and grip. A role-play tests service handoffs.
The prototype only needs enough fidelity for the question.
High fidelity too early can waste effort and make teams emotionally attached to one solution.
19. Fidelity should rise with commitment
Early prototypes should be cheap to discard.
Later prototypes can test material, performance and edge cases more realistically.
The cost of change rises through the design process.
Learning should therefore be front-loaded where possible.
20. A prototype is not proof of production readiness
One successful demonstration does not establish reliability, manufacturability or scale.
Prototypes often receive special attention and expert handling.
Production systems need repeatability under ordinary use.
Engineering Design owns the deeper route from concept into verified engineering.
21. Usability testing observes real interaction
Ask a participant to complete a representative task.
Observe where they hesitate, misunderstand or create workarounds.
A designer explaining how the interface is supposed to work contaminates the test.
The interface should carry enough meaning itself.
22. Success rate and completion time reveal different things
Original example: eight of ten users complete a task successfully.
Task success rate is 80 per cent.
If successful users take between 20 seconds and six minutes, the average hides large variability.
Design evaluation should inspect both completion and friction.
23. Errors are evidence about the interface
If several users make the same error, the problem may be design rather than user carelessness.
Labels, hierarchy or system state may be unclear.
A mature designer asks why the interface invited the mistake.
Repeated user error is often a design signal.
24. Edge cases reveal hidden assumptions
What happens with a very long name? No internet connection? One-handed use? A screen reader? An empty dataset?
Main-path testing proves only the main path.
Design quality often appears at boundaries.
Edge cases should be chosen by likelihood and consequence.
25. Iteration is evidence-driven revision
Iteration does not mean endless tinkering.
It means using test results to change the next version.
The loop is prototype → observe → diagnose → change → retest.
Without diagnosis, iteration can polish symptoms while preserving the underlying problem.
26. Visual design creates hierarchy
Size, contrast, spacing, alignment and colour guide attention.
A page can contain all necessary information and still fail because priorities are visually equal.
Hierarchy should reflect task importance.
Decoration is useful when it supports, rather than competes with, meaning.
27. Typography controls reading conditions
Type size, line length, weight and spacing affect legibility.
Typography also carries tone.
Designers should not trade readability for stylistic novelty when comprehension is the job.
The best typography often becomes invisible because reading feels effortless.
28. Colour should not carry the only meaning
Red and green status indicators can fail for colour-vision differences or low-quality displays.
Pair colour with labels, symbols or position.
This creates redundant representation.
Accessibility often improves general robustness.
29. Grids create consistent relationships
A grid aligns content across a page or screen.
Consistency reduces the amount users need to relearn.
Grids should support content rather than force every element into identical proportions.
Visual systems work when repetition helps prediction.
30. Interaction design makes system state understandable
Users need to know what the system can do, what it is doing and what happened after an action.
A button press with no feedback creates uncertainty.
A loading indicator, confirmation message or visible state change closes the loop.
Good interaction design preserves the relationship between action and consequence.
31. Affordances and signifiers guide action
An affordance is what an object or interface allows.
A signifier communicates where or how action is possible.
A flat text label that looks like ordinary prose may technically be clickable but poorly signify interaction.
Usability depends on what users can perceive, not what developers know exists.
32. Defaults are design decisions with population effects
Many users accept default settings.
A default therefore has more influence than a hidden optional preference.
Designers should choose defaults carefully when privacy, money or safety is involved.
A default can simplify use without becoming manipulation.
33. Error messages should support recovery
“Error 417” names the system’s internal state.
“Payment could not be confirmed. Your card has not been charged. Try again or use another method” supports the user’s next action.
Good error design explains consequence and recovery.
Failure states deserve as much design attention as success states.
34. Service design coordinates visible and invisible work
A customer sees the frontstage: website, counter, call centre, delivery.
Behind it sits backstage work: databases, approvals, logistics, policies and staff.
A service blueprint maps both.
This reveals why a front-end problem can originate in a backstage constraint.
35. Queueing is part of service experience
Waiting time affects perceived quality.
Design can reduce actual waiting or improve information during the wait.
A truthful estimate can feel better than unexplained uncertainty.
Service design therefore includes time and expectation.
36. Handoffs are fragile design boundaries
When one department transfers a customer or case to another, information can be lost.
Design should specify what state travels with the handoff.
The user should not become the courier for information the organisation already possesses appropriately.
Cross-System Handoffs owns the general mechanism.
37. Accessibility is a design quality, not an afterthought
People vary in vision, hearing, mobility, dexterity, cognition and language.
Design that assumes one ideal user excludes people unnecessarily.
Accessible design considers multiple ways to perceive, understand and operate a system.
Retrofitting after release is usually harder than designing access from the beginning.
38. Inclusive design uses exclusion as evidence
If one person cannot use a product because the interaction requires two hands, the barrier may affect many others temporarily or situationally.
A parent holding a child, an injured user and a person carrying bags can experience a similar constraint.
Designing for edge users can therefore improve mainstream use.
Accessibility expands the system’s operating envelope.
39. Universal design aims for broad usability without unnecessary special routes
A ramp can support wheelchair users, prams and deliveries.
Captions support deaf users and people in noisy environments.
Clear language supports cognitive accessibility and second-language users.
Universal design looks for shared solutions while recognising that some specialised accommodations remain necessary.
40. Sustainable design follows lifecycle consequence
A product has impacts before and after use.
Materials are extracted, processed, transported, maintained and eventually reused, recycled or discarded.
A lightweight product can have high environmental cost if it is replaced frequently.
Design therefore considers lifecycle, not only purchase moment.
41. Repairability preserves product value
Products can fail because one small component wears out.
If that component is inaccessible or unavailable, the whole product may be discarded.
Repairable design considers fasteners, diagnostics, parts, documentation and skill.
Maintenance is part of design.
42. Circular design keeps materials in useful loops
Reuse, remanufacture and recycling require materials to remain identifiable and separable.
Joining incompatible materials permanently can make recovery difficult.
Circularity therefore begins at design rather than at the recycling bin.
The Materials Engineering edition owns deeper material behaviour.
43. Systemic design widens the boundary beyond one artefact
The Design Council’s Systemic Design Framework extends divergent–convergent design thinking into complex social and environmental systems, adding attention to relationships, leadership and continuing the journey. Source: Design Council.
This matters because changing one service can move problems elsewhere.
Systemic design asks how interventions alter the surrounding network.
44. Design changes behaviour and therefore carries ethical responsibility
Interfaces can make actions easy, hard, visible or hidden.
That power can support users or manipulate them.
Designers should examine consent, privacy, safety, fairness and dependency.
“The user clicked it” is not enough when the interface was engineered to confuse.
45. Dark patterns are designs that exploit users
Examples include hidden fees, disguised advertisements, obstructive cancellation or false urgency.
These techniques can improve short-term conversion while reducing autonomy and trust.
Ethical design makes material consequences understandable before commitment.
The user should know what action they are taking.
46. Persuasive design should preserve agency
A health app can remind users to exercise.
A learning app can prompt retrieval practice.
Persuasion becomes problematic when it hides incentives or makes refusal intentionally difficult.
The design objective should be behaviour aligned with the user’s legitimate goals.
47. Design systems create reusable consistency
A design system contains components, patterns, tokens and guidance used across products.
It reduces repeated decisions and interface drift.
A design system should not freeze innovation.
It stabilises common elements so designers can focus attention where new reasoning is needed.
48. Brand design and product design meet at expectation
Brand design sets visual and emotional expectations.
Product and service design determine whether experience fulfils them.
A polished identity cannot compensate for a frustrating product.
Marketing owns brand positioning; Design owns the tangible and experiential expression.
49. Design management connects design decisions to organisational strategy
Large organisations run many design projects simultaneously.
Design management coordinates standards, teams, priorities, research and quality.
It protects coherence without requiring every product to look identical.
The Management edition owns the broader organisational coordination layer.
50. Failure analysis: diagnose the design mechanism
| Observed problem | Useful design question |
|---|---|
| Users cannot find a feature | Is hierarchy, naming or navigation inconsistent with their mental model? |
| Prototype tests well but launch fails | Did production, scale or real environment introduce conditions the prototype omitted? |
| Service feels fragmented | Are backstage handoffs losing information or ownership? |
| Accessibility defects appear late | Was inclusive testing absent from discovery and development? |
| Team keeps polishing one concept | Was divergence closed before alternatives were explored? |
| Users make the same error repeatedly | What in the interface invites that action? |
51. Learning workshop with worked answers
Question A: three handoff locations, three notification methods and two identity methods. Basic combinations? Answer: 18.
Question B: eight of ten users complete a task. Success rate? Answer: 80 per cent.
Question C: users repeatedly click non-clickable text. Is training the only solution? Answer: no. The visual signifiers may be misleading.
Question D: a team begins by building the requested app before observing users. Which phase was skipped? Answer: sufficient discovery and problem definition.
Question E: a beautiful package cannot be opened by users with limited grip. What does this reveal? Answer: aesthetic success did not establish functional accessibility.
52. Common misconceptions
“Design means making things look good.” Visual design is one domain inside a larger problem-solving discipline.
“User-centred means users design everything.” Designers gather user evidence and then exercise professional judgement under broader constraints.
“The Double Diamond is a strict sequence.” The Design Council explicitly presents design as iterative rather than a fixed linear recipe.
“A prototype is nearly finished.” Prototype fidelity depends on the question being tested.
“Accessibility is a specialist add-on.” Accessibility is part of whether a design is usable by its intended population.
53. Working glossary
Design: disciplined creation and refinement of forms, interactions or systems to fit defined human and contextual needs. Discovery: investigation used to understand the problem and users. Framing: definition of what problem is being solved.
Divergence: deliberate expansion of alternatives. Convergence: selection and refinement among alternatives. Prototype: representation built to test a design question. Usability: effectiveness, efficiency and experience of use under stated conditions.
Service blueprint: map linking visible user experience with backstage processes. Design system: reusable patterns and standards supporting consistent products. Inclusive design: practice that uses human diversity and exclusion as evidence in design decisions.
54. Evidence and disciplinary boundary
The process framework follows the Design Council’s Double Diamond, its Framework for Innovation and Systemic Design Framework. This article deliberately remains separate from Architecture, Engineering Design and Marketing, which own different downstream jobs.
The deeper answer: design works by making assumptions cheap to discover before mistakes become expensive
The power of design is not that designers always know the answer.
It is that good design creates a disciplined route through not knowing: observe the world, question the frame, generate alternatives, make ideas tangible, test them with reality, learn quickly and increase commitment only as evidence improves.
Continue: Project Management carries selected solutions through delivery; Architecture owns built-space design; Marketing owns market value and positioning. Return to the How X Works Hub.