Laws of UX

Ten principles including Fitts's law on target size and distance, Hick's law on choice and decision time, Jakob's law on users' expectations formed elsewhere, Miller's law on working memory, the serial position effect, Postel's law, and the peak-end rule.

Each is given its research origin, its design consequence and examples of products applying it well and badly.

Each principle is given its actual research origin rather than being asserted as a design maxim.

Devmark’s reading — not a quotation from the book

The discipline of tying each rule to its source is the book's contribution. A great deal of design advice circulates as folklore, and Yablonski consistently shows where a claim came from and what it actually said.

Jakob's law — that users spend most of their time on other sites and expect yours to work the same way — is the one most often ignored and the one with the clearest commercial consequence.

It is short and deliberately not exhaustive; these are heuristics, not a theory of interaction, and Yablonski says so.

Some principles are stretched: Miller's seven plus or minus two is frequently over-applied in design writing, and while he is more careful than most, the chapter still invites the usual misuse.

Miller's law is the one the field routinely over-applies, and this book only partly resists that.

Devmark’s reading — not a quotation from the book

Yablonski's framing is that these are constraints to design within rather than rules to check compliance against, and the distinction matters. Hick's law does not mean fewer options is always better; it means that adding an option has a measurable cost that should be paid deliberately.

The book is at its most useful as a shared vocabulary for a team. A designer who can name the principle a decision is trading against can argue for it, which is harder when the argument is about taste.

Two of the principles have direct, measurable consequences. The time to reach a target depends on its size and distance, which is why primary actions are large and near where the hand already is. The time to choose grows with the number and complexity of options, which is why long undifferentiated menus slow people down more than the extra scrolling would suggest.

The most consequential principle for anyone designing something new: people spend the overwhelming majority of their time on other products, so they arrive expecting yours to work like those. Yablonski's point is that convention is not a failure of imagination but an inherited model that a novel interface must be good enough to justify overwriting.

Your users formed their expectations somewhere else. Almost all of their time is spent there.

Devmark’s reading — not a quotation from the book

He is careful with the most commonly abused finding in design. The research concerned the capacity of short-term memory for unrelated items and says nothing about how many navigation items a menu should have; the useful derivative is chunking — grouping related information — rather than a numeric limit on anything visible on screen.

A principle worth more attention than it usually gets: every system has an irreducible amount of complexity, and the only real question is who absorbs it. Simplifying an interface frequently means moving the difficulty onto the user or onto the engineering, and a design that appears to have removed it has usually relocated it.

The observation that a system responding fast enough changes behaviour — not merely satisfaction but the pace and confidence of use — supplies a concrete performance target that is a design constraint rather than an engineering nicety, and it is the principle most often ignored because it belongs to nobody's remit.

Experiences are recalled by their most intense moment and by how they ended rather than by the average of everything that happened, which means effort spread evenly across a long flow is largely wasted on parts nobody will remember. The practical consequence is to decide deliberately where the peak should be and what the last screen says, and to concentrate the design effort at those two points rather than distributing it.

The closing section notes that every principle in the book is equally usable to manipulate: the same findings about attention, choice and reward underwrite persuasive design and dark patterns. It is short, and its inclusion at all is more than most references of this kind manage.