Book summary “Don't Make Me Think”: summary

Introduction

You open a web page and immediately wonder where you are, where to start, and what to click. You encounter unmarked links, endless text, and unclear choices that force you to stop and think just to complete basic tasks.

Steve Krug addresses this friction in Don't Make Me Think, A Common Sense Approach to Web Usability. The central idea is that web pages should be self-evident and obvious, requiring no conscious effort to understand.

You will learn how to design intuitive navigation, create effective home pages, and run simple usability tests. The book answers several core questions about digital design:

• How do you make web pages self-explanatory?
• Why do people muddle through sites instead of reading carefully?
• How do you test usability on a small budget?

1. Guiding Principles of Usability

When you look at a web page, every question mark adds to your cognitive workload. It distracts you from the task at hand.

The overriding principle of usability is to eliminate unnecessary thought, making every page or screen self-evident and obvious.

Steve Krug calls this the First Law of Usability: don't make me think. If a page cannot be made completely self-evident, it must at least be self-explanatory. Understanding it should require very little effort.

An online booking site might fill your head with mental chatter and errors. Another site takes your inputs and gives logical choices with no question marks at all.

First Law of Usability

You do not read web pages carefully. Instead, you scan them, satisfice by choosing the first reasonable option, and muddle through without understanding how things work.

Because you are usually in a hurry, you act like a shark and keep moving. You focus only on words and phrases matching your current task or interests. You choose the first plausible path rather than optimizing. You use technology successfully without learning its underlying mechanics.

Gary Klein studied fire commanders who did not compare multiple options under high pressure. They took the first reasonable plan and tested it mentally.

You need to design web pages like highway billboards that can be grasped in a fraction of a second by users who are whizzing by. This approach requires Billboard Design 101 principles:

• rely on established conventions;
• establish clear visual hierarchies;
• break pages into defined areas;
• make clickable elements obvious;
• eliminate visual noise;
• format text to support rapid scanning.

Harlem.org built its site entirely around Art Kane's photo of fifty-seven jazz musicians as an innovative yet usable alternative to text links.

While consistency is helpful, clarity is always the superior goal when designing interfaces. Designers often invoke consistency as an absolute rule. If making something slightly inconsistent makes it significantly clearer to you, clarity should always be chosen.

How Users Really Browse

You do not mind clicking multiple times as long as each click is a mindless, unambiguous choice that keeps you on the right track. The number of clicks matters less than the amount of thought and uncertainty required for each click.

Robin Burgener created 20q.net, an online game that uses a neural net algorithm to play Twenty Questions.

Street corners throughout London feature "LOOK RIGHT" instructions painted on curbs to save tourists from unexpected traffic.

Most words on web pages are skipped by users. Removing half the words is a realistic and beneficial goal. Happy talk and introductory instructions must be removed or pruned to the bare minimum.

Pruning a 103-word survey introduction down to 34 words makes instructions actually usable.

Krug's Third Law of Usability requires you to follow specific rules:

• Get rid of half the words on each page.
• Get rid of half of what is left.

2. Navigation and Home Page Design

Web navigation serves as the foundational physical and conceptual structure of a site, substituting for the missing physical cues of the real world.

Unlike physical stores where users have a sense of scale, direction, and location, the Web lacks physical dimensions. Users feel inherently lost, making navigation the primary mechanism to embody the site hierarchy and create a sense of place.

Persistent Navigation

Persistent navigation must appear consistently on every page except forms, including four essential items: Site ID, Sections, Utilities, and Search.

Keeping navigation consistent confirms to users they remain on the same site and spares them from relearning how it works. The Site ID must be in the upper left corner to reflect its position at the top of the logical hierarchy.

The search function on a website should follow a strict formula consisting of a box, a button labeled Go, and the word Search or a magnifying glass icon. Users immediately scan for standard search patterns. Adding fancy wording like Find or Quick Search, instructions, or scope options into the persistent search box forces users to think unnecessarily.

Every page needs a prominent name that matches the link clicked, alongside clear current location indicators to prevent disorientation. An implicit social contract dictates that the page title must match the words clicked by the user. Furthermore, You are here indicators and Breadcrumbs must be made highly prominent because users routinely miss subtle visual cues.

The Trunk Test

Good Web navigation can be evaluated using the trunk test, verifying if a user dropped randomly onto any page can identify key orienting elements. If a page is well designed, a user with blurred vision should instantly recognize the Site ID, page name, sections, local navigation, current location indicators, and search mechanism.

You can run this test with a simple procedure:

• Choose a page anywhere in the site at random, and print it.
• Hold it at arm's length or squint so you can't really study it closely.
• As quickly as possible, try to find and circle each of these items: Site ID, Page name, Sections (Primary navigation), Local navigation, You are here indicator(s), Search.

The Big Bang Theory

The home page must immediately communicate the big picture, answering what the site is and why the user should use it within the first few seconds.

Home pages face intense stakeholder pressure and promotional overload, often causing the main purpose of the site to get lost. Users form snap judgments within milliseconds that guide their subsequent interpretations. An experiment timeframe measuring how quickly people form initial impressions of a Web page takes only 50 milliseconds.

Promoting too many features on the home page destroys its overall effectiveness through a classic tragedy of the commons. Because home page real estate is highly coveted, stakeholders constantly push to add promotions. While the promoted section gains traffic, the collective effectiveness of the home page is degraded by clutter for all users.

William Forster Lloyd originated the shared resource concept, later popularized by biologist Garrett Hardin in his 1968 Science essay The Tragedy of the Commons regarding overpopulation and town pastures.

3. Usability Testing on Ten Cents a Day

Usability testing resolves team arguments by shifting the focus from subjective opinions about what people like to objective observations of what works.

Web teams frequently waste time in religious debates driven by personal likes and professional perspectives, assuming that most users share their preferences. Individual reactions are based on too many variables for one-dimensional preferences to apply. Instead of arguing over abstract preferences, teams should build a crude version of a feature and watch real people attempt to use it. Testing defuses impasses and reveals the vast diversity of user motivations and behaviors.

Focus Groups versus Usability Tests

Focus groups and usability tests serve entirely different purposes and must not be confused with one another. Focus groups involve a small group of people discussing opinions, feelings, and abstract needs, making them useful in the planning stages to determine if an idea makes sense. Usability tests involve watching one person at a time actually use a site or prototype to perform typical tasks so you can detect and fix things that confuse or frustrate them.

Marketing teams often request focus groups near launch dates to avert potential disaster, but this fails to reveal whether the site actually works. A typical focus group involves 5 to 10 participants.

Running Monthly Sessions

Web development teams should commit to a fixed, simple testing schedule of one morning a month to ensure it actually gets done. Spending one morning a month keeps testing simple and affordable, provides enough actionable insights to keep the team busy fixing things, frees the team from tying tests to slippery project milestones, and increases attendance from stakeholders. Three participants are tested per monthly morning session.

Testing just three participants per round is ideal for do-it-yourself usability testing because qualitative insights matter more than statistical proof. Do-it-yourself testing aims to improve the product by identifying usability flaws rather than scientifically proving a hypothesis. Because you will always find more problems than you have resources to fix, three participants are more than enough to uncover the most serious issues.

Strictly recruiting users who precisely match a narrow target audience is often unnecessary and hinders frequent testing. Most sites suffer from usability flaws that cause problems for almost anyone. By loosening requirements, you save time and money, and you can make allowances for differences by asking whether a user's difficulty stems from their lack of domain knowledge or a genuine site flaw. Older recruiting guides apply a 20 percent inflation rate, while average web user incentives for a one-hour session run from 50 to 100 dollars.

To keep testing practical, you can follow specific steps:

• try to find users who reflect your audience without getting hung up on narrow profiles;
• make allowances for differences between participants and your actual audience;
• ask yourself if a problem occurred only because the participant lacked specialized knowledge.

Encouraging team members, managers, and executives to observe usability tests transforms their perspective on users. Watching users struggle with a site is a transformative experience that helps stakeholders realize that users do not think the same way they do. Providing an observation room with screen sharing and good snacks helps lure people in to watch.

Chocolate croissants are specifically noted as working particularly well to lure observers in.

An effective observation space requires specific arrangements:

• set up an observation room with a computer, screen sharing software, and a large screen monitor or projector;
• have observers write down the three most serious usability problems during the break after each test session.

A standard one-hour usability test session follows a strict, repeatable structure ranging from introduction to final probing. The session begins by putting the user at ease, explaining that the site is being tested rather than the user, and getting permission to record. It moves through background questions, a home page tour without clicking, specific tasks while thinking aloud, and concludes with probing questions and wrapping up.

Steve Krug walks an imaginary 25-year-old router named Janice through a test session for eLance.

The test session follows a precise timeline:

• welcome and explain how the test will work taking 4 minutes;
• ask background questions about occupation and internet usage taking 2 minutes;
• conduct a home page tour asking the participant to look around and narrate taking 3 minutes;
• watch the participant perform tasks while thinking aloud taking 35 minutes;
• ask probing questions about what happened taking 5 minutes;
• thank the participant, pay them, and show them to the door taking 5 minutes.

Teams must debrief immediately after tests to collectively identify and ruthlessly focus on fixing the most serious problems first. By reviewing observed problems over lunch right after testing, the team can avoid getting bogged down by low-hanging fruit or new feature requests, ignoring minor stumbles that users quickly recover from on their own. The top 10 serious problems are selected and rated during debriefing.

To handle the results systematically, apply these rules:

• make a collective list of observed problems by going around the room sharing notes;
• choose the ten most serious problems;
• rate and number them from 1 to 10 with the worst at the top;
• create an ordered list detailing how each will be fixed in the next month;
• stop once all available time and resources are allocated.

Remote and unmoderated testing methods offer alternative ways to gather user feedback without requiring in-person office visits. Remote testing lets participants use screen sharing from their own homes or offices, expanding the recruiting pool beyond local geographic limits. Unmoderated remote testing services let you submit tasks and receive recorded videos of users completing them within an hour, with an average turnaround of one hour for unmoderated remote testing videos.

Services like UserTesting.com provide recorded videos of users testing tasks and thinking aloud.

You can execute unmoderated sessions through straightforward steps:

• send your tasks and a link to your site, prototype, or mobile app to an unmoderated testing service;
• watch the resulting video of a user completing your tasks.

4. Mobile Usability, Goodwill, and Accessibility

Designing for mobile devices involves navigating extreme constraints like tiny screens and unreliable connections, requiring careful tradeoffs rather than hidden features.

Steve Krug explains that mobile screens force sites to make hard tradeoffs. Many mobile usability failures stem from poor compromise decisions, such as breaking stories into tiny, slow-loading multi-page chunks with repeated ads or photos.

Steve Krug stops using CBS News on his iPhone because its stories are broken into too-small chunks that take too long to load and force users to scroll past the same photo repeatedly across 8 pages.

Mobile Constraints

Because creating separate mobile and desktop site versions leads to out-of-sync content and double work, developers rely on scalable design. To assist users before dynamic layouts are perfected, sites should follow three key suggestions.

• Allow zooming on your site for mobile devices.
• Ensure links from emails or social media take users directly to the requested article rather than the mobile Home page.
• Always provide a link to the full Web site toggle at the bottom of every page.

Because fingers touching glass cannot be accurately sensed until contact is made, mobile devices lack a cursor. This eliminates hover actions, tool tips, changing button states, and drop-down menus, forcing designers to compensate using remaining visual dimensions.

Unlike desktop testing, mirroring mobile touch screens hides the user's physical gestures and taps, making sessions hard to follow. Krug recommends attaching a lightweight webcam to the device using a book light clip and a gooseneck clamp.

Steve Krug built a portable mobile testing camera rig called the Brundleyfly by merging a book light clip with a webcam, costing about 30 dollars in parts and taking an hour to build.

You can execute testing using simple tools:

• Connect the Brundleyfly camera to the facilitator’s laptop via USB.
• Open a display view window like AmCap on a PC or QuickTime Player on a Mac for the facilitator to watch.
• Share the laptop screen with observers using screen sharing software like GoToMeeting or WebEx.
• Run a screen recorder like Camtasia on the observation room computer.

Reservoir of Goodwill

Users enter websites with a limited reservoir of goodwill that is depleted by friction and poor decisions, but it can be refilled by considerate design.

Every usability problem, hidden phone number, or forced data-formatting rule lowers the user's goodwill reservoir. Conversely, sites can replenish goodwill by anticipating user needs, being transparent about costs and issues, and saving users steps.

Steve Krug describes checking an airline's website during an impending strike, finding zero mention of the strike anywhere on the site, and experiencing a total depletion of goodwill toward the airline.

Sites hurt user trust by hiding customer support numbers, obscuring prices, and making users format data like credit card numbers or phone numbers manually. To restore trust, sites must make main tasks obvious, be upfront about fees, and provide easy error recovery.

• Know the main tasks users want to do and make them obvious and easy.
• Tell users what they want to know upfront, including shipping costs, parking fees, and service outages.
• Save users steps wherever possible, such as embedding direct tracking links in email receipts.
• Maintain up-to-date, candid Frequently Asked Questions lists based on actual customer support queries.
• Provide creature comforts like printer-friendly pages.
• Make it easy to recover from errors, and apologize when user requests cannot be met.

Accessibility Steps

Developers often dismiss accessibility statistics or doubt that adaptations benefit everyone. However, the true argument for accessibility is that it dramatically alters lives, such as allowing blind individuals to independently read newspapers through screen readers.

Automated validators often return vague warnings, but real progress comes from fixing general confusion, observing screen reader users, studying dedicated accessibility books, and applying low-hanging technical fixes.

• Fix usability problems that confuse everyone so users with disabilities do not face compounded barriers.
• Read Mary Theofanos and Janice Redish's article observing blind users working with screen readers.
• Read a book on web accessibility such as A Web for Everyone by Sarah Horton and Whitney Quesenbery or Web Accessibility by Jim Thatcher et al.
• Go for the low-hanging fruit by adding appropriate alt text to images, using HTML heading elements correctly, making forms work with screen readers with label elements, putting a Skip to Main Content link at the top of each page, ensuring keyboard accessibility, maintaining strong text-to-background contrast, and using an accessible template.

10 Key Ideas

1. The overriding principle of usability is to eliminate unnecessary thought, making every page or screen self-evident and obvious.
2. Because users act like sharks and scan pages in a hurry, designers must follow Billboard Design 101 principles by using visual hierarchies, established conventions, and pruned text.
3. Persistent navigation must appear consistently on every page except forms, incorporating standard search patterns and clear current location indicators.
4. The home page must immediately communicate the big picture by answering what the site is and why the user should use it within the first few seconds.
5. Usability testing resolves team arguments by shifting the focus from subjective opinions about what people like to objective observations of what works.
6. Web development teams should commit to a fixed, simple testing schedule of one morning a month, testing just three participants per round to uncover the most serious issues.
7. Designing for mobile devices involves navigating extreme constraints like tiny screens and unreliable connections, requiring careful tradeoffs rather than hidden features.
8. Users enter websites with a limited reservoir of goodwill that is depleted by friction and poor decisions, but it can be refilled by considerate design.
9. Sites must maintain user trust by making main tasks obvious, being upfront about fees and customer support, and providing easy error recovery.
10. The true argument for accessibility is that it dramatically alters lives, while technical progress comes from fixing general confusion, observing screen reader users, and applying low-hanging technical fixes.