Don't Make Me Think

Review

I read Don't Make Me Think for the first time shortly after I started working in product, and it is one of the few books I would recommend to someone on their first day. Not because it is revolutionary, but because it is precise. Krug names what every working designer knows implicitly but rarely says out loud, and he does it without wasting your time.

The observation that landed hardest on a first read: users scan, they do not read. They do not deliberate, they satisfice. They do not figure things out, they muddle through. If you design for an attentive, deliberate user, you are designing for someone who does not exist. Getting that wrong is responsible for more bad products than almost anything else.

The usability testing section is the most practical part of the book. The case for testing early, cheaply, and repeatedly is made convincingly without requiring a UX research background to execute on. The book is clear that the timing of testing matters more than the scale of it.

The web has changed substantially since 2000. The principles have not. Worth reading early in a product career and worth revisiting when you forget why obviously simple things turn out to be so hard to get right.

Key Takeaways

The parts worth keeping:

Testing

• Most usability debates are a waste of time. Teams defend strongly held positions that cannot be proven. Nobody changes their mind. Build a version and watch three people try to use it. One afternoon of testing ends more arguments than weeks of discussion, and teaches you that user reactions are varied and unpredictable.
• Test before you think you are ready. Getting someone in front of a rough prototype early, even a competitor's product, reveals more than waiting until you feel the design is presentable. Timing matters more than polish.
• Keep testing lightweight so you actually do it. Three to four people per round will surface most of the problems. The valuable rhythm is: test, find the worst issues, fix them, then run another round. Repeated small rounds outperform a single large study by a significant margin.
• Recruit loosely and grade on a curve. Representative users matter less than people think. Experts are not insulted by a product that is clear enough for beginners. Offer an incentive, keep the invitation brief, and do not discuss the site beforehand.
• Use testing to inform your judgment, not to prove or disprove a predetermined position.
• You will find problems faster than you can fix them. Triage ruthlessly: ignore issues where users recover quickly, take feature requests with a grain of salt, and focus on the most serious problems first. When fixing, resist the impulse to add things. The best answer is often to remove something.
• Every unsolved problem drains the goodwill reservoir. Users accumulate frustrations. When the reservoir empties, they leave and do not come back.
• Accessibility is one of the few places where a small amount of effort can dramatically improve someone's experience. Fix the usability problems that confuse everyone first.

How users actually behave

• Users do not read web pages. They scan. They glance at a page, spot something that resembles what they want, and click it. They do not evaluate options before choosing. They pick the first reasonable link. Design for the behavior you actually see, not the behavior you wish they had.
• There is no big penalty for guessing wrong, so users satisfice: they choose good enough, not best. Your job is to make the good-enough choice feel obvious so they do not even have to think about it.
• Users do not figure out how things work. They muddle through. If your audience is going to act like they are reading billboards, design great billboards.
• Eliminate question marks. Every unanswered question adds cognitive load and diverts attention from the task: where do I start, can I click that, why did they call it that, where am I. Each one erodes confidence a little further.
• There is a continuum from obvious to obscure. Almost always skew toward obvious. When users have to puzzle over how something works, you have already eroded their enthusiasm and trust in the product.
• Usability means something works well. Pay attention to it and you reduce frustration, increase satisfaction, and increase the chance users come back.

Navigation

• Browsing a website is like moving through physical space, except without the natural cues of scale, direction, and location. Navigation is what replaces those cues. Standardize its appearance and differentiate it clearly from page content.
• Drop onto any page on a well-designed site and you should be able to answer six questions immediately. Krug calls this the trunk test.

• Every page needs either a search box or a link to a search page.
• The home page carries more load than any other page: site identity, hierarchy, search, the best content, timely content, shortcuts, and registration. Everyone in your company will want a piece of it. Do not let that win.
• Home pages attract short-sighted behavior. Each promotion individually looks reasonable but they compound. The page ends up clamoring for attention everywhere and overall effectiveness drops. Protect three things above all: where to search, where to browse, and where to find the best content.

Making pages work

• Every page should be self-evident: users should instantly understand what it is and how to use it without spending any effort thinking about it. If self-evident is not achievable, at minimum make it self-explanatory.
• Five tools for pages that work at a glance: visual hierarchy (important things are prominent, related things look related, nesting shows containment); conventions (copy what works and reinvent only when you genuinely have something better); clearly defined page zones so users can ignore irrelevant areas; obvious clickability; minimal noise. Assume everything is visual noise until proven otherwise.
• Click count is the wrong metric. What matters is the amount of thought required and the level of uncertainty at each step. Users do not mind multiple clicks if each one feels confident and clear. What kills usability is the moment of doubt. Jared Spool calls this the scent of information.
• Making choices mindless is one of the primary levers for improving usability.
• Omit needless words. Half the words on most pages can be cut without losing anything. Fewer words reduce noise and make the content that remains more prominent.