GBDA 101: Foundations of Digital Media Design

Sources and References

Primary textbook — Robin Williams, The Non-Designer’s Design Book, 4th ed. (Peachpit Press). Supplementary texts — Ellen Lupton Thinking with Type; Lupton & Phillips Graphic Design: The New Basics; Don Norman The Design of Everyday Things; Don Norman Emotional Design; Tim Brown Change by Design; Josef Albers Interaction of Color; Philip B. Meggs Meggs’ History of Graphic Design. Online resources — IDEO design thinking toolkit; Stanford d.school open materials; Adobe Illustrator and Photoshop official tutorials.

Chapter 1 — Design Thinking and the Design Process

Design is not decoration. It is a way of approaching problems whose solutions cannot be deduced from a formula, whose success depends on how people feel and behave, and whose answers are negotiated between competing constraints. Tim Brown, in Change by Design, frames design thinking as the discipline that matches people’s needs with what is technologically feasible and commercially viable. Three words are worth underlining: desirability, feasibility, and viability. The interesting design problems live where all three overlap.

The IDEO toolkit and the Stanford d.school describe the same loop in different languages. IDEO talks about inspiration, ideation, and implementation. The d.school talks about empathize, define, ideate, prototype, test. The labels matter less than the rhythm: move out into the world to learn about people, sharpen the question, generate more possibilities than seem reasonable, make cheap artefacts that let you think with your hands, and put them in front of real users so reality can argue back. Each loop should be short, and you should run many.

Empathy is the part students skip. It is tempting to read a brief, form an opinion, and sprint toward an aesthetic. The method instead asks you to watch, interview, and shadow. Listen to what people say, watch what they actually do, and pay attention to the friction between the two. A shop sign people walk past without seeing is a signal. A piece of software where users open three menus to reach a common feature is a signal.

Defining the problem is a deliberate act of narrowing. A good problem statement names the user, their need, and the insight you discovered — “a first-year student needs a way to find quiet study space on short notice because they feel socially exposed asking strangers if a seat is free.” The insight is the interesting part. Without it, the statement has no leverage and the ideas that follow feel generic.

Ideation assumes the reliable way to get a good idea is to have many and then edit. Brainstorms work best when framed by a specific “How might we” question, run under a clock, and protected from premature judgement. Volume, wildness, building on others’ ideas, and deferring critique are the classic rules because they counteract the pathologies that kill group creativity.

Prototyping converts ideas into things you can touch. The prototype is not a miniature of the final product but a question you are asking. A paper sketch asks whether the general layout makes sense. A clickable mockup asks whether navigation feels natural. A full visual design asks whether the brand reads as intended. Fidelity should match the question — polishing a button on a prototype whose core flow is still wrong is wasted time.

Testing is where the designer becomes a student again. The purpose of a user test is to discover how the design fails. You give people a task, watch them attempt it, and resist the urge to explain. Their confusions, mis-clicks, and off-hand remarks are the data. You then return to an earlier stage and run another loop.

Chapter 2 — The Sketch as Tool of Thought

Sketching is to design what writing is to thinking: a way of externalising ideas so they can be examined and improved. A sketch is not a drawing. A drawing represents something faithfully; a sketch makes an idea visible quickly enough that you can have the next idea while you are still interested. The virtue of the sketch is its cheapness — because each takes seconds, you are willing to throw it away, and because you are willing to throw it away, you are willing to explore a bad idea to its end and learn from it.

Lupton and Phillips in Graphic Design: The New Basics argue that sketching is how designers build a vocabulary — a private library of shapes, marks, and relationships that they can later deploy deliberately. The practical exercise most courses use is the grid of thumbnails: draw a rectangle, divide it into eight or sixteen cells, and force yourself to produce one variation per cell. The first four are obvious, the next four uncomfortable, and the last eight surprising. You are not trying to make good compositions; you are exhausting the easy ones so the harder ideas have room to appear.

Tools do not matter much. A pencil and paper are enough, and many professionals prefer them because the friction of a physical medium slows them down just enough to think. The deepest benefit of sketching is what it does to your attention. Because a sketch commits so little, you look at what is in front of you rather than what you believe should be there. A beginner who sketches twenty logos sees more about logos in an hour than one who opens Illustrator and starts placing vector shapes.

Chapter 3 — Visual Perception and Gestalt Principles

Design operates on vision, and vision is not a neutral camera. The human visual system aggressively organises incoming light into objects, groups, and figures, and it does so according to predictable rules. Early twentieth-century Gestalt psychologists catalogued these rules, and designers have been using them ever since, whether or not they know the names.

Proximity says that elements placed close together are read as belonging to the same group, and elements placed far apart are read as separate. Robin Williams makes proximity the first of her four principles in The Non-Designer’s Design Book because almost every beginner mistake in layout is a proximity failure: the caption drifts from its image, every block of text gets the same amount of air around it, related controls end up on opposite sides of a dialog. The fix is nearly always to pull related items closer and to push unrelated items apart.

Similarity says that elements sharing a visual attribute — colour, shape, size, orientation — are read as belonging to the same category, even when they are not close to each other. This is what lets a table of contents work: every chapter title looks like every other chapter title, so the eye groups them as a set.

Continuity says that the eye follows implied lines and smooth curves. A column of text rides its baseline grid because the baselines form a continuous line, and the same principle lets designers lead a reader’s eye across a complicated layout with nothing more than the alignment of a few edges.

Closure says that the eye completes shapes that are only partially drawn — an idea famously exploited in logos such as the IBM stripes and the World Wildlife Fund panda. Figure and ground describes the relationship between the shape you are attending to and the space behind it; good logos work in both directions, using negative space as an active element.

Common fate — elements moving or changing together are perceived as one group — matters in interactive work, where motion binds a cluster of controls even if they look different. These rules are not arbitrary aesthetic preferences. They are defaults of the visual system, which means you can either work with them or pay a penalty in confusion.

Chapter 4 — Colour Theory for Designers

Colour theory is both simpler and stranger than beginners expect. Simpler, because only a handful of relationships account for most well-composed palettes. Stranger, because, as Josef Albers argued in Interaction of Color, a colour never exists on its own; its appearance is shaped by whatever sits next to it. Two identical patches of grey look cooler or warmer depending on the surround. The same red looks lively against black and muddy against orange. The practical consequence is that you cannot pick colours in isolation — you must always see them in the context in which they will appear.

The colour wheel is a convenience, not a physical law, and there are several competing versions of it. The additive wheel (red, green, blue) describes how light combines on a screen and is what software uses under the hood. The subtractive wheel (cyan, magenta, yellow plus black for printing) describes how inks absorb light. The traditional artist’s wheel (red, yellow, blue) is the one in most introductory books and the one used to describe harmonies. For digital design work you will meet all three, and your main job is to remember which one governs which medium.

The practical vocabulary of colour has three axes. Hue is the colour’s identity — red versus green versus blue. Saturation, sometimes called chroma, is how pure or grey the colour is. Value, or lightness, is how close to white or black it lies. Confusing value with hue is the most common beginner mistake. Two colours that have identical value will not create contrast, and a layout that relies on them to separate elements will look flat and hard to read. When in doubt, convert your design to greyscale; anything that reads there will read in full colour, and anything that disappears there needs help.

Harmonies are predictable relationships around the wheel. A monochromatic palette uses one hue at various values and saturations; it is the safest choice and rarely wrong. Analogous palettes use neighbours on the wheel and feel calm and natural. Complementary palettes use opposites and feel energetic because the opposite hues enhance each other’s saturation at the boundary. Split complements, triads, and tetrads extend the idea. None of these are laws — they are starting points for the eye.

Colour also carries meaning, and the meanings are partly cultural and partly physiological. Warm colours advance, cool colours recede; red alarms, blue calms; yellow reads as cheap or as friendly depending on saturation and context. Cultural connotations vary, and a palette that signals celebration in one country can signal mourning in another. A designer working for a global audience checks these assumptions deliberately.

Accessibility is not optional. A significant fraction of any audience has some form of colour vision deficiency, and many more read your work on cheap screens in poor lighting. WCAG specifies contrast ratios — 4.5:1 for body text, 3:1 for large text — and any system that ignores them will be illegible for part of its audience. Never rely on colour alone to convey information; pair it with shape, icon, or label.

Chapter 5 — Composition and the Principles of Design

Williams’s The Non-Designer’s Design Book organises layout around four principles that she calls by the acronym CRAP: contrast, repetition, alignment, proximity. They are not the only principles, but they are a complete starter kit, and most layout failures are a failure of one of them.

Contrast is the principle that if two elements are not the same, they should be obviously different. Near-misses — a headline that is only slightly larger than the body, two greys that are almost but not quite identical, two fonts that are both sans-serif and about the same weight — look like mistakes. Contrast can be created through size, weight, colour, value, font, direction, or shape. When you commit to a contrast, commit hard.

Repetition means taking some visual element from your design and repeating it throughout the piece. Repeated elements knit a composition together and make multi-page documents feel like a set. The repetition can be obvious (a recurring rule line at the top of every page) or subtle (a typeface used only for pull quotes everywhere), but it should be consistent enough that the reader learns the pattern and is surprised when it breaks.

Alignment is the principle that nothing on a page should be arbitrary. Every element should have a visual connection to something else on the page. Beginners tend to centre everything, and centred layouts can work, but they leave soft, fuzzy edges that weaken the composition. Left or right alignment produces the crisp invisible lines that a reader’s eye follows automatically. A grid — columns, gutters, and baselines — is an alignment system made explicit.

Proximity, already introduced under Gestalt, is the fourth principle: group related items and separate unrelated ones with space. Together these four tools cover most of what a layout needs.

Beyond CRAP, Lupton and Phillips introduce a broader vocabulary: balance, hierarchy, rhythm, emphasis, unity. Balance can be symmetrical, as in a classical poster, or asymmetrical, as in a modern magazine spread where a heavy photograph on one side is counter-weighted by white space and small type on the other. Hierarchy is the explicit ordering of importance — what should the eye see first, second, third? You make hierarchy visible through scale, weight, colour, placement, and isolation. Rhythm is the pattern of repeats and intervals in a composition; emphasis is the deliberate breaking of that pattern to call attention; unity is the overall sense that every part belongs to the same piece. A designer who can name these moves can critique their own work, which is the prerequisite for improving it.

Chapter 6 — Typography Fundamentals

Typography is design’s most demanding subject because most people, including many designers, underestimate it. Ellen Lupton’s Thinking with Type begins with the claim that letters are not an add-on to design but one of its central materials. A page of text is an image, and the image is made out of the shapes of the letters, the spaces between them, and the rhythm of lines down the column. Learning to see type is learning to see a whole new dimension of the page.

The anatomy of a letter is worth the small investment of memorising. Every letter has a baseline, an x-height (the height of the lowercase body), a cap height, ascenders rising above x-height and descenders dropping below the baseline. It may have serifs — the small strokes at the ends of main strokes — or not. The aperture is the opening in letters like c and e; counters are the enclosed spaces inside letters like o and d. These details are what distinguish typefaces from one another.

Typefaces divide into broad historical families. Old-style serifs (Garamond, Caslon) derive from Renaissance calligraphy and have moderate contrast between thick and thin strokes. Transitional and modern serifs (Baskerville, Bodoni) have sharper contrast and more vertical stress. Slab serifs (Clarendon, Rockwell) have thick, square terminals. Sans-serifs divide further into grotesque (Helvetica, Akzidenz-Grotesk), humanist (Gill Sans, Frutiger), and geometric (Futura, Avenir) categories. Scripts imitate handwriting. Display faces are designed for headlines, not body text, and usually fail at small sizes. Choosing a typeface well starts with understanding what family you need.

The measurable choices that govern how type reads on the page are called its micro-typography. Type size is the most obvious and the one beginners over-tune — for body text, something between 9 and 12 points in print and 14 to 18 pixels on screen usually works. Line length matters almost as much — a comfortable line holds roughly 45 to 75 characters, and a line much longer than that makes the reader’s eye lose the start of the next line. Leading, the space between lines, should grow as lines get longer; too tight and lines crowd each other, too loose and the text falls apart into stripes. Tracking and kerning — the space between letters — are usually left at the designer’s defaults for body text and adjusted by hand for large display settings.

Hierarchy in typography is built out of contrasts: size, weight, style (roman versus italic), case, colour, and spacing. You do not need many levels. Most documents work with three or four — a title, a section heading, a subheading, and body text, perhaps with a caption style for margins. A common beginner mistake is to give each hierarchy level a different typeface; the more durable move is to pick one or two families and get variety from weight and size within them.

Pairing typefaces is intimidating and mostly unnecessary. A single well-designed family with several weights — say, a humanist sans like Source Sans Pro, or a classic serif like Adobe Garamond — will carry almost any document. If you do pair, the reliable combinations are a serif body with a sans headline, or a humanist sans body with a condensed display for titles. Pair by contrast (they should be obviously different) but also by voice (they should feel like they belong to the same project). Avoid pairing two typefaces that are similar but not identical; those are the ones that look like mistakes.

Finally, typography lives on grids. A well-set page has a baseline grid that holds the lines of body text, a column grid that holds the text blocks, and a margin system that frames the whole. On the web the grid is flexible but the principles are the same. Grids are freedom, not constraint — they solve a hundred questions in advance so that you can spend your attention on the one question that matters for this particular layout.

Chapter 7 — A Brief History of Graphic Design

Philip Meggs wrote Meggs’ History of Graphic Design to argue that graphic design, though it became a profession only in the twentieth century, has a history as long as writing. The history matters to a designer the way grammar matters to a writer — not so that you recite it, but so that your choices are informed by the voices that came before you.

The earliest chapters belong to the invention of writing — cuneiform tablets, Egyptian hieroglyphs, the Phoenician alphabet that seeded Greek and eventually Latin letters. The Roman square capitals carved into Trajan’s Column are still the reference for proportion in classical typefaces two thousand years later. Medieval scribes in monastic scriptoria developed blackletter and uncial scripts and the illuminated manuscript tradition that turned books into objects of devotion. The arrival of Gutenberg’s movable type in the 1450s made type into an engineering problem and put books into the hands of the middle class, changing everything from religion to politics.

The Industrial Revolution introduced mechanised printing, lithography, photography, and advertising — and with them the first explicitly commercial graphic design. Victorian posters were loud, layered, and ornament-heavy. The Arts and Crafts movement, led by William Morris, reacted against that visual noise by returning to handcraft values and carefully designed type. Art Nouveau followed with organic curves and integrated lettering, and then the modernist movements of the early twentieth century — Futurism, Dada, De Stijl, Constructivism, the Bauhaus — rewrote the rules.

The Bauhaus in particular matters because so much of current digital design still lives in its aftermath. Walter Gropius and his colleagues argued for functional simplicity, geometric clarity, and the integration of art, craft, and industry. Bauhaus typography favoured sans-serif faces and asymmetric grid-based layouts. After the Nazi government closed the school in 1933, its teachers scattered to the United States and other countries, carrying those ideas into post-war design education.

Swiss Style — sometimes called the International Typographic Style — emerged after the Second World War and cemented the modernist playbook: strict grids, sans-serif typefaces (Helvetica, released in 1957, being the most famous), generous white space, and photography over illustration. Designers like Josef Müller-Brockmann and Armin Hofmann taught a whole generation that clarity was the highest virtue. The style became the default visual language of corporate identity for decades.

The late twentieth century reacted against Swiss neatness. Postmodern designers like Wolfgang Weingart and April Greiman broke grids, layered type, and used the computer’s new possibilities to create deliberately messy compositions. Meanwhile, the arrival of the web forced designers to think about screens, interactivity, and audiences who would never touch the printed page.

Digital design is heir to all of this. When you open Figma today, your defaults — the grid, the sans-serif interface type, the clean whitespace — are modernist inheritance. Knowing the history makes your choices choices rather than accidents.

Chapter 8 — Icon Design: Symbols that Speak

An icon is a small image that has to do a great deal of work. It must be recognisable at a glance, legible at sizes from sixteen pixels to hundreds, distinct from its neighbours, and consistent with any other icons in its set. Designing icons well is harder than it looks, and it rewards the slow, subtractive process that beginners tend to avoid.

The first decision is semantic. What does the icon stand for? The easy cases are nouns with obvious referents — a house for home, a magnifying glass for search, an envelope for mail. The hard cases are abstract concepts — an icon for “trust” or “process” or “synchronise” has no physical referent and must rely on convention. The designer’s job in those cases is to borrow an established metaphor rather than invent a new one, because an icon the user has to learn from scratch is a broken icon.

The second decision is style. Is the set line-based or filled? Geometric or organic? Sharp-cornered or rounded? Mono-weight or varying? These choices should be made once at the beginning of a set and enforced across every icon. Consistency is what separates a professional set from a pile of clip art.

The third decision is structural. Every icon in a set should share the same grid, the same visual weight, and the same optical size. Optical size is subtle: a circle and a square of identical bounding-box size will appear to be different sizes because the circle contains less area. Icon designers compensate by growing circles slightly beyond the square grid so that they look the same size. Similarly, a horizontal rectangle and a vertical one at the same pixel size will feel differently weighted, and the designer has to adjust by eye.

Drawing the icon itself is usually done in a vector program — Adobe Illustrator, Figma, or the free alternative Inkscape — because vectors scale without loss. Start with simple geometric primitives (circles, rectangles, triangles), combine them with boolean operations, and refine until the shape reads at the smallest intended size. Test at 16, 24, and 48 pixels on the screen. If it does not read at 16, it does not work.

Pixel grid alignment matters at small sizes. A stroke of 1.5 pixels renders as a blurry two-pixel stroke because displays cannot draw half-pixels. Icon designers snap every line to whole pixels at the target size, sometimes producing different versions for different sizes. At very small sizes you may need to drop detail entirely.

Finally, test the icon in context. An icon that is beautiful in isolation but unrecognisable in a crowded toolbar has failed. Place your icons next to each other, next to labels, next to real UI backgrounds, and ask whether a user who had never seen them could guess what they mean.

Chapter 9 — Visual Identity, Branding, and Logo Design

A logo is not a brand. The brand is the set of associations that someone holds in their head when they think of an organisation, and it is built out of everything that organisation does — its products, its words, its service, its visual presentation. The logo is only the most compact visual signal for that brand. Understanding the difference saves a designer from the common trap of trying to make a logo do more than a logo can do.

A good logo is simple, distinctive, appropriate, memorable, and scalable. Simple because it has to survive reproduction at every size and in every medium, from a favicon to the side of a bus. Distinctive because it needs to separate its organisation from its competitors in a crowded field. Appropriate because the style, colour, and feel should match what the organisation actually does — a bank logo and a children’s toy logo should not be interchangeable. Memorable because the logo’s main job is recognition. Scalable because the logo will be used in places the designer cannot predict.

Logos divide into a few categories. Wordmarks are typographic treatments of the organisation’s name — Google, Coca-Cola, FedEx. Lettermarks use initials — IBM, CNN, HBO. Pictorial marks are recognisable images — Apple’s apple, Twitter’s bird. Abstract marks use non-representational shapes — Nike’s swoosh, the Chase octagon. Combination marks mix a symbol and a wordmark. Emblems embed the name inside a shape, like a badge — Starbucks, many sports teams. Each category has its uses, and the choice should follow the brief, not the designer’s current fashion.

The process of designing a logo looks a lot like the design-thinking loop in miniature. You begin with research — what does the organisation actually do, what is their voice, who are their customers, what do competitors look like? You sketch widely, filling pages with thumbnails before you go near the computer. You narrow to a handful of directions and refine each one in vector form. You test them at tiny sizes, in single colours, and against the backgrounds where they will appear. You present two or three options with rationales, not thirty. You iterate based on feedback and build out the usage guidelines once the mark is settled.

A visual identity system extends the logo into everything else. The system specifies the primary and secondary typefaces, the colour palette with exact values, how much white space must surround the logo, what background colours are allowed, how the logo should be cropped or animated, and how photography, illustration, iconography, and voice should feel. These rules are collected in a brand guidelines document sometimes called a brand book. The point of the book is not to constrain designers but to ensure that a customer who sees a business card, a website, and an app from the same organisation experiences them as the same organisation.

Branding fails most often through inconsistency. A clean logo applied messily across ten touchpoints creates a weaker brand than a plain logo applied consistently. The designer’s long-term job, once the identity is built, is to maintain it — a discipline Williams’s repetition principle serves at the scale of a whole company.

Chapter 10 — Emotional Design

Don Norman’s The Design of Everyday Things famously argued that most objects are frustrating because their designers did not think about the people who would use them. His later book, Emotional Design, takes that argument further. He claims that usability alone is not enough — that the emotional response a product creates is not a decoration on top of utility but a functional part of the design. People who enjoy an object use it differently, forgive its flaws more readily, and even solve problems more creatively while using it. Aesthetic and affective qualities are therefore as important to a product’s success as its feature list.

Norman proposes three levels of emotional processing, each of which a designer can target. The visceral level is immediate, pre-conscious, and physical. It is the gut reaction you have to a shape, colour, texture, or sound before your brain has named the object. Visceral design is the domain of sensory appeal — the glossy curve of a well-made tool, the reassuring heft of a phone, the particular red of a Coca-Cola can. Good visceral design is not the same as prettiness; it is the match between a surface and the feelings the designer wants the viewer to have in the first quarter second.

The behavioural level is the realm of use. A product that feels good at the visceral level can still be a misery if it obstructs the tasks a user is trying to accomplish. Behavioural design is classical usability: affordances, feedback, consistency, error tolerance. Does the button look pressable? Does pressing it produce visible acknowledgement? Are similar actions in similar situations handled similarly? Can the user recover from a mistake without losing their work? These are the questions Norman has been asking since the first edition of The Design of Everyday Things, and the answers still determine whether a product is actually pleasant to use.

The reflective level is the slowest and the most cultural. It is the story the user tells themselves about owning and using the product — what it says about them, how it fits into their identity, how it will be remembered. Reflective design is the domain of brand, meaning, and status. A watch can tell the same time as a cheaper one and still matter differently to its owner because of what it symbolises. Reflective design is where craft meets narrative, and designers who ignore it tend to produce work that is fine in the moment and forgettable afterward.

The three levels are not independent. A visceral appeal that the behavioural level disappoints creates frustration — a beautiful gadget you cannot figure out. Behavioural competence with no visceral charm produces respect without affection. Reflective meaning without behavioural competence is marketing without product. The designer’s job is to consider all three deliberately.

The practical implication for digital media is that micro-details matter. The timing of an animation, the sound of a notification, the particular curve of a loading spinner, the wording of an error message — each of these is an emotional cue. None of them shows up on a feature list, but together they determine whether a user feels respected or annoyed, whether they return or churn. Learning to pay attention to emotional details is a lifelong exercise, and it starts with asking the question Norman asks on every page: how does this make the person feel?

Chapter 11 — Designing Interactions

An interface is a conversation between a person and a system, and like any conversation it can be clear or garbled, respectful or rude. Steve Krug’s Don’t Make Me Think collapses the principles of good interaction design into a single rule: the user should not have to stop and puzzle over what to do next. The rule is harder to apply than it sounds, because what seems obvious to the designer usually is not obvious to someone seeing the screen for the first time.

The classical vocabulary Norman gives us is affordance, signifier, mapping, feedback, and constraint. An affordance is a relationship between an object and a user that makes a certain action possible — a button affords pressing, a door handle affords pulling. A signifier is the visible cue that communicates the affordance — the shadow under the button, the shape of the handle. The designer’s job is not only to put the affordance there but to make sure the signifier is loud enough that users notice it. A button that looks flat and unclickable has the same usability problem as a door with no visible handle.

Mapping is the relationship between controls and their effects. Good mapping is natural — the up arrow scrolls up, the volume slider on the right controls the right speaker, the lights in a room correspond spatially to the switches on the wall. Bad mapping is any relationship where the user has to learn an arbitrary correspondence. Feedback is the system’s acknowledgement that the user’s action has been registered — a button that briefly highlights, a form that shows a saving state, a loading spinner that tells you something is happening. The absence of feedback is how users decide a system is broken and start clicking things twice.

Constraints prevent errors by limiting what the user can do in a given state. A greyed-out button that cannot be pressed until required fields are filled is a constraint. A date picker that prevents picking impossible dates is a constraint. Constraints should be visible and understandable; an error message that appears only after the user has tried and failed is worse than a constraint that prevented the failure in the first place.

Krug’s book adds several more practical rules. Users do not read pages — they scan them, so layouts should be built around visual hierarchy, short paragraphs, bulleted lists, and headings that actually describe content. Conventions are friends; a navigation bar at the top of a page, a logo that returns home, a magnifying glass that means search — these are shared vocabularies your users already know, and you should only break them when you have a very good reason. The home page should answer, within a second of loading, the question “what is this?” and “what can I do here?” If your users cannot answer those questions quickly, no amount of visual polish will save the site.

Accessibility overlays all of interaction design. A button that is too small for a thumb, a contrast ratio too low for a user with low vision, a flow that cannot be completed with a keyboard alone, a label that screen readers cannot parse — each of these excludes someone. The Web Content Accessibility Guidelines provide a concrete checklist: text alternatives for images, keyboard navigability, sufficient contrast, meaningful sequence, predictable behaviour. Designing for accessibility is not a separate mode you switch on at the end; it is a set of constraints you carry throughout.

Chapter 12 — Prototyping: Low Fidelity to High Fidelity

A prototype is a question you can hold in your hand or see on your screen. The purpose of the prototype is to let you test that question — with yourself, with collaborators, with real users — without committing to the cost of a finished product. The art of prototyping is matching the fidelity of the artefact to the question you are asking, and resisting the temptation to polish before the underlying ideas have been tested.

Low-fidelity prototypes are cheap, fast, and disposable. A paper prototype is literally drawings on paper — a set of screens you swap in and out as the tester “taps” through them. Paper prototyping works startlingly well for testing basic navigation, information architecture, and flow, because the crudeness of the medium encourages honest feedback. Testers feel comfortable telling you the layout is wrong when they can see that you spent thirty seconds drawing it; they are much less willing to argue with a pixel-perfect mockup.

Medium-fidelity prototypes are usually digital wireframes. Tools like Figma, Sketch, or Adobe XD let you draw screens out of boxes and placeholder text, link them with clickable hotspots, and share them with reviewers. Wireframes answer questions about layout, labelling, and flow. They deliberately strip away colour and typography so that the conversation stays on structure. Beginners sometimes resist this stripping because they want to see the design “for real,” but the reason to strip it out is precisely to avoid arguments about hue when you should be talking about hierarchy.

High-fidelity prototypes bring in the final visual treatment — actual colours, fonts, photography, and sometimes real content. They answer questions about emotional tone, brand fit, and details that only become visible at full resolution. The cost is that changes grow more expensive as fidelity grows — a button colour that is one click to change in Figma can become a week of work once it is baked into code.

The most common mistake is jumping to high fidelity too early. You spend three days perfecting a screen whose core flow has not been tested, and the test then reveals that users need the screen to do something different. The discipline is to treat each fidelity level as a separate question and to move up only when the lower layer is settled. Prototypes are also about communication — when you hand a stakeholder an interactive mockup, they see what you mean in a way that a spec document never conveys.

Chapter 13 — User Testing and Iteration

User testing is the moment when your design meets reality, and reality has a lot to say. The purpose of the test is not to prove that the design works. The purpose is to discover the ways it does not, so that the next iteration can be better. A test that “went well” in the sense that nothing went wrong is usually a test that did not learn anything — either because the tasks were too easy, the sample was too friendly, or the observer was leading the participant.

Krug argues that user testing is radically cheaper and more useful than most teams believe. You do not need a research lab, a statistically significant sample, or a month of preparation. Three or four participants, a working prototype, a list of tasks, and an hour of observation will surface most of the major usability problems in a design. Run the test monthly, and you will outpace teams that plan elaborate tests and never actually run them.

The mechanics are straightforward. Recruit participants who roughly match your target users — not your co-workers, and especially not the designers who made the thing. Explain to them that you are testing the design, not them, and that anything they do will be useful. Give them a realistic task — “imagine you are looking for a quiet place to study tonight; please find one using this app” — and then sit back and watch. Ask them to think aloud so that you can hear what they are puzzling over. Resist the urge to explain, defend, or help. When they get stuck, note what they tried and what they expected.

What you are looking for are moments where the user’s model of the system diverges from yours. They click something you did not expect to be clickable. They ignore a button you thought was obvious. They use the wrong word for the thing they are trying to do. They say “I thought I could just…” These are the data points. They should not be taken as commands — users are often wrong about what they want — but they are reliable as signals that something in the design needs examination.

After the test, collect what you learned and decide what to do. Not every problem is worth fixing, and not every fix will work on the first try. Prioritise: serious confusions that most testers hit get fixed first, idiosyncratic preferences get parked, and anything you are unsure about goes back into the next round of testing. Then iterate. The speed of iteration is more important than the depth of any single pass. A team that runs ten small tests in a month will learn more than a team that runs one huge test at the end of the quarter.

Iteration is the discipline of staying a student longer than is comfortable. It is tempting to fall in love with a design and defend it; the practice is to treat every version as a question and every test as a chance to be surprised.

Chapter 14 — Critique: Giving and Receiving Design Feedback

Design is a social craft. Even solo designers lean on mentors to catch blind spots, and most work is made in teams where critique is constant. Critique is a skill that has to be learned, because the naive form of feedback — liking or not liking something — is almost useless.

Useful critique begins with observation. Before saying whether a design is good, describe what you see. “The headline is the largest element. The photograph bleeds off the left edge. The secondary text is in muted grey.” Description grounds the conversation in the artefact instead of in taste, and it often reveals to the designer things they did not realise they had made.

Interpretation comes next. What is the design trying to communicate? What hierarchy does it imply? What tone does it strike? This is where critic and designer compare their models, and where the most interesting disagreements happen. A design that was meant to feel playful but reads as chaotic has a gap between intention and effect, and the gap is what the designer needs to know.

Judgement — liking or not liking — is the last and least useful stage, and it should be expressed in terms of the design’s own goals rather than the critic’s preferences. “This feels too formal for the audience we discussed” is a useful judgement. “I just don’t like the colour” is not, unless you can explain why the colour fails the brief.

Receiving critique is its own skill. The instinct to defend your work is natural and counterproductive. Listen first, take notes, and resist the urge to explain every choice. If a viewer misread the hierarchy, the answer is not that they misread it but that you wrote it poorly. Write down even the suggestions you disagree with; sometimes the value is in the underlying observation that the design is not landing.

The most productive cultures treat critique as a normal, low-stakes event rather than a showdown. Structured formats like “I like, I wish, what if” make it easy to give honest feedback without anyone feeling attacked. Critique is also how you learn to critique yourself, which is the ultimate form of the skill — catching problems before anyone else has to. That is what it means to become a mature designer: not to need less feedback, but to use every piece of feedback more completely.

Closing Note

Design is a practice, not a body of knowledge to memorise. These chapters give you a vocabulary — Gestalt principles, colour harmonies, CRAP, typographic hierarchy, Norman’s three levels, Krug’s interaction rules, the loop of prototype and test — but vocabulary is only the beginning. The work happens when you sit down with a real brief, sketch twenty ideas you do not love, narrow to three, build the cheapest possible version of one, put it in front of a stranger, watch them struggle, and start over without despair. Do that loop enough times and you become a designer.