Every system has irreducible complexity. The only question is whether the user absorbs it or the product does.
I'm designing
A Checkout
The last few screens, where every unnecessary decision costs a sale.
What this surface turns on
- Whether an account is required before payment.
- The order of steps, and whether it matches how people think about buying.
- How address and card entry tolerate messy input.
- What the confirmation actually says.
Where these usually go wrong
- Forced account creation before payment.
- Card numbers rejected because the user typed the spaces printed on the card.
- Address fields that impose one country's structure on everyone.
- Generic payment errors that leave the user guessing which detail was wrong.
- Entered data cleared on failure.
Why these
Checkout is the highest-stakes form most products have. These entries cover cost transparency, payment failure, guest access and the confirmation, because those are the four places where money is actually lost.
Why it works the way it does
The reasoning layer. Read these when a decision needs justifying, or when something is failing and you cannot say why.
Be liberal in what you accept and conservative in what you send, which for interfaces means accepting messy input and returning clean output.
Effort increases as a visible goal gets nearer, and artificial early progress accelerates the whole journey.
People judge an experience by its most intense moment and its ending, not by the average of the whole thing.
Working memory holds a small number of chunks, so the design lever is what counts as a chunk, not the raw item count.
Losing something feels worse than gaining the same thing feels good, which shapes how people react to change.
The craft decisions
The visual and structural foundations that carry most of the weight on this surface.
Things placed close together are read as belonging together, before anything is read.
A fixed set of spacing values, so every gap in the product comes from the same vocabulary.
Using size, weight, colour and space to make the structure of text visible before it is read.
Difference is what makes anything perceivable, and it is measurable for text against its background.
Making one thing stand out, which only works if most things do not.
Patterns to reach for
Solutions people already recognise, with the conditions under which each one earns its place.
Breaking a long task into steps, which reduces per-screen load but adds navigation cost and abandonment points.
Checking input at the right moment, which is rarely while the user is still typing.
Suggestions as the user types, which turns recall into recognition when it is fast enough.
An interruption that should be reserved for genuinely irreversible actions, because it stops being read when it is routine.
What to show while waiting, chosen by how long the wait is and whether its length is known.
Components you will specify
Each entry covers the states, behaviour and accessibility contract you need to define.
Free-form entry, where the label, the hint and the error placement do most of the work.
Choosing one option from a known set, which is the wrong control below about five options.
Triggers an action. If it navigates somewhere, it should be a link instead.
A persistent message tied to a context, which is where anything requiring action belongs.
Communicating how far through something the user is, which requires honesty about what you can measure.
Accessibility requirements
Not a separate pass. These are the requirements this surface raises, mapped to the decisions that produce them.
Errors must be identified in text, described specifically, and accompanied by a suggested fix where one is known.
Every input needs a persistent label, and format requirements must be stated before the user types.
Everything must be operable with a keyboard alone, which is the single most load-bearing accessibility requirement.
Deliberately moving focus when the interface changes, so keyboard and screen reader users are not stranded.
Interactive targets need to be large enough and spaced enough to hit reliably.
Language and content
The words do more work here than most teams budget for.
The small pieces of text that do most of the work: labels, hints, empty states, errors and confirmations.
Say what happened, why if you know, and what to do next, in that order and in plain language.
Designing so the product can be adapted to other languages and regions without being rebuilt.
How to find out
Research methods suited to the questions this surface raises.
Measuring progression through a sequence, where the largest drop is rarely the biggest opportunity.
Watching people attempt real tasks, which is the most direct evidence available about whether a design works.
Comparing variants with real traffic, which tells you what happened without telling you why.
Check your work
Interactive checklists, with progress saved in your browser.
Product pages, cart and checkout, covering the decisions that determine whether people buy.
A review pass for any form, covering structure, labelling, validation, errors and accessibility.
Whether the design serves the user's interest, including where that now overlaps with legal obligation.
Documents you might need
Fillable templates, prefilled so they are usable rather than empty.
Every path through a task, including the error paths and the states most flows forget.
A session script with tasks phrased as goals, and moderator prompts that do not lead.
Decisions other teams made
Worked cases, including the options that were rejected and why.
A retailer required registration before purchase to build its customer database, and lost the sales it was trying to capture data about.
Splitting a long form into steps did not raise completion until the team also saved progress and let people come back.
Designing something else?
Neighbouring territory. Each surface reuses much of the same material, arranged around the decisions that surface turns on.