My friend took it to work at her web design firm, and it kept disappearing off her desk.

For example, think of your average stereo or mp3 player. If you wanted to hear music, what would you do? Just about everyone would scan the object with their eyes, trying to find a symbol or word to guide them. In this case, they'd look for a right pointing triangle, commonly known as a "play" button. This all happens within milliseconds, subconsciously, but we all do it. The easier it is to find what you're looking for and engage that option, the higher usability marks the object receives. That's a rather simplistic explanation of usability, but it's the basis of this book when applied to the web.

Author Steve Krug makes a living critiquing web sites to establish how easy they are for your average Joe to use. He gathers willing participants, sits them down in front of a computer, and asks seemingly obvious questions like "do you know where you are?", "do you know what this company does?", "where would you click first, and why?", etc.

It's not rocket science, that's for sure. But if a company is planning on making money off of their web presence, a confusing web site could put them in the dog house with consumers.

And that's the important thing to realize about "Don't Make Me Think" -- this book is written from the viewpoint of what "works" with e-commerce sites. It's the greasing of the wheels to make a shopping experience fast, intuitive, trusting, and most importantly, EASY.

If you are a graphic designer, consider the web to be your personal playground to create, engage, shatter conventions and develop new ideas in design, this book is not for you.

Krug acknowledges this up front, but insists that the lessons learned in this book can be applied to just about ANY web design that deviates from the norm. And in a sense, that's quite true. If you were to build a shining new city on the hill with ultra modern buildings and cutting edge style, you'd still need an effective transportation system to tie your creation together.

But where the book stumbles is that Krug only promotes (or acknowledges) one ubiquitous web design style -- a company logo in the upper left corner, folder tab navigation across the top, and subsections down the left side. Sound familiar? It should...take a look at the top of this page.

P>His opinions would be more trustworthy if he would critique non-standard, but successful examples of web design that are just as effective. But he only offers one solution, and it's one all of us are very aware of already.

Krug does have useful points to remember -- including consistency of style, navigation, and brand identification -- but when you get down to it, it's all COMMON SENSE. Open your browser, check out eBay, Outpost.com, or Salon.com, and study their navigational systems, use of color and contrast, and copy length. See how they allow you to move between pages, subsections, and the placement of navigational icons. Then ask yourself, how did I KNOW to click there? Was it the wording? The color?

This book would be a lifesaver to anyone who was just starting out designing for audiences on the web however, so if you have yet to design your first web site, by all means pick this up. The lessons to learn here are just as important (if not more so) as learning HTML. Because if you didn't want people to use and enjoy your work, you wouldn't be putting it on the web.

Bottom line: I learned how to conduct beautifully effective user testing with minimal investment, time, and reading!

1. Don't Make Me Think!

The number one usability rule, most often expresed by users. Web pages should be self-evident, obvious, and self-explanatory. Buttons should have short text and look clickable. The default search for your site should be simple.

2. Design for scanning not reading

By observing users Krug found that people glance, scan some text, and click on the first reasonable option (called "satisficing"). People scan Web pages, they don't read them. We don't make optimal choices, we satisfice.

Here are some things you can do to make sure users understand as much of your site as possible:

a. Create a clear visual hierarchy to show relative importance of content (H1/H2 etc.)
b. Take advantage of conventions
c. Break pages up into clearly defined areas
d. Make it obvious what's clickable
e. Minimize noise

3. Users like mindless choices

Make each click an unambiguous orthogonal alternative.

4. Omit needless words

Get rid of half of the words on each page, then get rid of half of what's left. This is especially important on home pages and
gateway pages.

5. Navigation: Use street signs and breadcrumbs

Factoid: The back button accounts for 30 to 40 percent of all Web clicks. Persistent navigation appears on every page of the site and should include the following five elements:

a. Site ID
b. A way home
c. Search
d. Sections
e. Utilities

Your navigation should answer these questions:

a. What site is this?
b. What page am I on?
c. What are the major sections of this site?
d. What are my options at this level?
e. Where am my in the scheme of things?
f. How can I search?

6. Your home page should convey the big picture

What is the site about? Use a good short tag line and welcome blurb. Rotate site promotions. Remove everything nonessential.

7. Most Web design usability arguments are waste of time

These "religious debates" consist of people expressing strongly held personal beliefs about things that can't be proven. All Web users are unique. There are no average users. There are no simple "right" answers for most Web design questions. What works is good integrated design that fills a need, that's carefully thought out, well executed, and tested.

The antidote for religious debate is to ask specific questions and test with real users. The last three chapters of the book show how to perform testing on the cheap with three or four users. I really enjoyed this book, especially Krug's easy humor. From WebReference.com.

For one thing, Steve Krug is funny. No kidding; you'll squirt coffee out of your nose on every page, at least if you're foolish enough not to put the cup down after page one.

For another, his usability principles are very straightforward, easily understood and remembered, and well illustrated by copious examples. Although this book will be most useful to people who actually design websites, it _could_ be read by pretty much anybody (including -- hint, hint -- the _bosses_ of people who actually design websites).

He's somewhat at odds with Jakob Nielsen on the question of usability testing, but that's okay. In the first place, such disagreement is healthy in a field where, really, nobody _knows_ anything with full certitude (even when people _say_ they do); and in the second place, Krug is telling us how to do usability testing on a shoestring, within real-world economic constraints. So I say: get Nielsen's _Designing Web Usability_ too, but get this one first.

Krug's book is well done all around. And if your boss happened to see it sitting on your desk, maybe even borrow it, the world of the Web might become just a wee bit more civilized.