Steve Krug's Don't Make Me Think is most famous for its argument about cognitive load, but it also describes something less often quoted: the specific patterns by which products erode user trust, slowly and usually without anyone on the team noticing.

Goodwill, in Krug's framing, is the reservoir of patience users arrive with - and certain design decisions drain it faster than others.

Hiding valuable information

Users arrive in a product with questions: where is the thing I'm looking for, how do I complete this task, what happens if I do this?

Every moment they spend looking for answers is a moment of friction, and friction has a compounding effect; a small amount is annoying, a consistent amount is alienating, and a product that reliably fails to answer basic questions will lose users to competitors who don't.

The solution isn't simply "make things more visible," though that's often part of it. It's to apply the principles of visibility, mapping, and feedback as a filter on every design decision - asking, at each stage, whether the user has the information they need to proceed with confidence.

Hidden information is rarely a deliberate choice; it's usually the residue of designing from the inside out, from the perspective of someone who already knows how the product works.

Punishing users for doing things the wrong way

People make mistakes; this is not a flaw in human cognition but a feature of it.

The question for designers is not how to prevent all error - that's neither possible nor desirable - but how to design for recovery. A user who deletes something important and immediately finds an undo function leaves the interaction having learned that the product looks after them; a user who has no path to recovery leaves having learned the opposite.

The design principle here is symmetrical: make errors hard to make in the first place, through clear affordances and well-considered interaction patterns; and make them easy to recover from when they occur, through undo functions, confirmation steps before irreversible actions, and error messages that explain what happened and what to do next.

Punishment by design - creating situations where a small mistake has large and permanent consequences - is one of the fastest ways to destroy confidence in a product.

Asking for information you don't need

Data collection is one of the defining ethical tensions of contemporary product design. Users are increasingly aware that what they share will be used in ways that extend beyond the immediate interaction, and that awareness shapes how they respond to data requests.

An onboarding flow that asks for a phone number when no feature requires it, a survey that appears before the user has had a chance to form a view, a tracking mechanism that runs without disclosure - each of these sends a signal that the product's interests and the user's interests are not aligned.

The test is simple: for every piece of information requested or collected, ask whether removing it would meaningfully harm the user's experience. If the answer is no, it shouldn't be there.

This isn't purely an ethical position; users who feel a product is collecting from them rather than serving them are less likely to engage, less likely to convert, and more likely to leave.

Making things inconsistent

Consistency is how users build a mental model of a product. When an element behaves the same way in every context - when a button always does the same kind of thing, when colour is used reliably to signal the same meaning, when spacing consistently indicates the same relationships - the user doesn't have to relearn the product as they move through it.

When those patterns break, the cognitive cost increases with each new inconsistency encountered.

The most common consistency failures are invisible to the designer because they accumulate gradually: a second team introduces slightly different components, a new feature is built without referencing the design system, a copy style drifts between screens.

The solution is structural: a design system, maintained and applied consistently, removes the cognitive load from individual decisions and creates the conditions for users to develop reliable expectations. The detail level matters too; users notice small variations in stroke weight, colour value, and spacing even when they couldn't articulate what felt off.

Designing without users

There is a particular kind of failure that produces beautiful products that barely work.

It comes from treating the visual and the structural as the primary challenge and treating actual user behaviour as something to be discovered at launch. Research doesn't require a large budget or a dedicated UX team; it requires a commitment to finding out whether the assumptions built into a design are accurate before those assumptions are shipped.

Competitive analysis, user interviews, usability testing with five participants, watching someone navigate a prototype for twenty minutes without intervening - all of these are forms of proximity to the user that change what gets built.

The designer who has watched three people struggle with the same interaction knows something that cannot easily be accessed from inside their own expertise, no matter how developed that expertise is.

Neglecting visual design

Visual craft signals intent. A product that looks unpolished communicates, at some level, that the people who built it didn't care enough to finish it - and that perception, driven by the aesthetic usability effect, shapes how users interpret everything else they encounter.

Users extend more patience to products that look considered, and less to products that don't; this is not a rational response, but it is a consistent and well-documented one.

The risk in overcorrecting is treating visual polish as a substitute for functional quality rather than a complement to it. A product can be visually refined and functionally clear; in fact, the disciplines reinforce each other, because the same attention to detail that produces good visual craft tends to produce careful interaction decisions too.

The goal is balance: neither sacrificing the experience for the aesthetics, nor treating visual quality as peripheral to the real work.

What runs through all six of these failures is a shared root: they occur when design decisions are made from the perspective of the product rather than the perspective of the person using it.

Goodwill is not earned through grand gestures; it is the accumulated effect of a product consistently demonstrating, through small decisions made well, that it was built for the person sitting in front of it.