By: Steve Krug
- Chapter 1. "Don't make me think"
- Chapter 2. "How we really use the web"
- Chapter 3. "Billboard design 101"
- Number 1 rule of web usability: Don't make users think
- Web pages should be self-evident.
- A user should immediately grok whatever page they land on
- Names, labels etc. should be obvious.
Jobs>Employment Opportunities>Job-o-rama - Links and buttons should be clearly clickable
- Names, labels etc. should be obvious.
- Users shouldn't spend time thinking about questions such as:
- Where am I?
- Where should I begin?
- Where did they put ____ ?
- What are the most important things on this page?
- Why did they call it that?
- If you can't make a page self-evident make it self-explanatory. This is the case when building something novel which requires designing outside of established templates/patterns.
- Motivation:
- Making pages self-evident improves a user's impression of the whole product
- Most users spend very little time looking at the pages we design. Self-evident pages are the most effective.
- Expectation: Users will examine each page, reading, and trying to understand it
- Reality: Users glance at each page. They scan quickly to find what they've come for
- Facts about web use
- #1 We don't read pages. We scan them.
- We're usually in a hurry.
- We don't care about most stuff on a page. We're looking for specific things of interest.
- We scan for words & phrases that match our current task or our general interests.
- #2 We don't make optimal choices. We satisfice.
- We don't choose the best option. We choose the first reasonable option we come across.
- We're usually in a hurry. Optimizing is time consuming.
- There's not much of a penalty for guessing wrong. (This seems like a principle for designing. i.e. as a designer/developer make sure that if users make an incorrect choice, they can always quickly and easily back out of it.)
- #3 We don't figure out how things work. We muddle through.
- Most users don't read instructions and they don't form a mental model of how the software works. They come up with a vaguely plausible working model that allows them to get stuff done.
- It's not important to users how software works.
- If we find a solution we don't deviate in the future. i.e. we don't look for better solutions.
- #1 We don't read pages. We scan them.
- If you can design something that users just "get" then:
- There's a better chance they'll use the full range of what you offer
- There's a better chance they'll be successful
- You can steer them better
- They'll feel better about themselves
- Create a clear visual hierarchy on each page
- Clearly portrary the relationships between things on the page
- Prominance is proportional to importance. Use size, color, white-space etc. to make more important things more prominent
- Visual relation indicates logical relations. Group related items together.
- Things are nested visually to show what is part of what.
- Conventions are your friend.
- Don't re-invent the wheel
- Conventions are useful. Users already understand them and it produces a sense of familiarity in users.
- Designers are often reluctant to use conventions. For the most part, resist the urge to innovate. Use conventions.
- If you are going to innovate make sure that the innovation is (1) completely clear and (2) adds so much value that the learning curve is justified.
- Break up pages into clearly defined areas
- This allows users to quickly scan the page to decide what section to focus on
- Make it obvious what's clickable
- Keep the noise down to a dull roar
- Noise makes a page hard grasp
- Busy pages are distracting. When everything on the page is aiming for you attention it's hard to focus
- Background noise is unhelpful. Assume everything is visual noise until proven otherwise.