UX elements and the three golden rules

◈ 4 cards

The five planes of user-experience work, the four cross-cutting concerns a software engineer owns, and the three golden rules with the sub-principles that carry the marks.

An interface is a set of decisions, not a residue

Most bad interfaces were never designed. They accumulated — one screen per feature, one control per option, defaults chosen by whoever wrote the field first. User experience (UX) design is the discipline that refuses that accumulation. Its working rule is that nothing reaches the release candidate unless someone decided it should be there, in that place, in that state, for a named user doing a named task.

That rule is easy to state and hard to apply, so UX work is split into planes that run from the most abstract commitment to the most concrete one. Strategy fixes who the users are and what the business needs from them. Scope turns that into a feature set and the content those features need. Structure decides how the system responds to a user action (interaction design) and how the content is organised (information architecture). Skeleton places things: information design, interface design, navigation design. Surface is the finished visual design. The ladder is useful because it tells you which argument you are actually having. A dispute about button colour is a surface dispute; a dispute about whether a member may hold two tools at once is a scope dispute, and no amount of visual polish will settle it.

The four concerns a software engineer owns

Of all this work, four cross-cutting concerns land squarely on the engineering team. Information architecture structures content for organisation, labelling, navigation and search. User interaction design covers the interface between product and person — what a tap, a gesture, a voice command or a scan can do, what feedback each produces, and which constraints prevent errors. Usability engineering specifies, designs and tests the human-computer interaction part of the product, so that vague praise like 'user friendly' is replaced by things you can measure: easy to learn, easy to use, easy to remember. Visual design is the look and feel; if nobody on the team has the eye for it, hire someone who does.

The three golden rules

Mandel's three rules are the spine of every interface argument you will be asked to make. Each expands into named sub-principles, and the sub-principles are where the marks are.

Place the user in control. Do not force the user into an interaction mode they cannot leave. Offer flexible interaction, because different people prefer different input. Make sequences interruptible and every action undoable. Let interaction streamline itself as skill grows, and let it be customised. Hide technical internals — the user should never meet the file system. Design for direct manipulation of the objects on screen.

Reduce the user's memory load. Every remembered item is a chance to make an error. Prefer recognition to recall by showing past actions rather than demanding them. Establish meaningful defaults, with a reset. Define shortcuts whose mnemonics are obvious. Base the layout on a real-world metaphor the user already understands. Disclose information progressively — high level first, detail on request.

Make the interface consistent. Give the user a visible sense of context — where they are, where they came from, what they can do next. Keep the rules the same across a whole product line. And do not break an interaction convention users already rely on unless there is a compelling reason.

The thinking move

When you are asked to justify an interface decision, do not say it 'feels cleaner'. Name the rule, name its sub-principle, then name the user cost the decision removes. Three sentences, and the answer stops being taste.

Worked example — the BorrowBox collection screen

A member is standing at the locker bank on a wet Tuesday evening with a bag in one hand. The first draft of the collection screen asked them to type the six-character reservation code printed in the confirmation email. Run the rules over it. Rule 2 fails immediately: the code is pure short-term memory load and the email is in another app. Replace it with a list of the member's active holds, one tap each — recognition instead of recall. Rule 1 fails next: the screen had no way back once a locker door was triggered, so a mis-tap cost a whole reservation. Add an undo window before the door opens. Rule 3 fails last: the librarian's return screen put the confirm button on the left while the member's screen put it on the right. One product, one position.

Each of those three changes is defensible in a sentence, and none of them is about colour.

Read the stack bottom-up: every concrete decision on the surface should be traceable to a strategy decision about who the user is and what the business needs. The type size of the BorrowBox locker number is a surface decision; whether a member may hold two tools at once is a scope decision, and arguing about it in a colour review wastes everybody.

The middle column is what an examiner is actually counting. A candidate who names the three rules and stops is answering about a third of the question; a candidate who names two sub-principles per rule and attaches each to a decision has written the whole answer.