Copyright: 2000
Publisher: Macmillan
ISBN: 0789723107

Don't Make Me Think: A Common Sense Approach to Web Usability is the result of Steve Krug sharing his expertise in a very controversial field.  If you count up all the web designers in the world, that number would come close to the number of opinions on the One Right Way to design a web site.  Krug addresses how to resolve these "religious debates" as well as how to take a common sense approach to web usability.  He also details a great way to do some usability testing which he insists is an absolute must if you are going to design a great site.

Question Marks

Krug asks us to imagine a user perusing our site for the first time.  Imagine thought clouds or bubbles popping up over the user's head as he or she pokes around the site.  Those thought clouds could include things like "Ah, that button allows me to search" or "There is the main menu" or "If I wanted to see the privacy policy I could click there."  These are good thought clouds but what if they thought clouds were "Where would I click if I wanted to see the purchase options?" or "I wonder if that button will search just this site of the entire internet?"  In these cases, the thought clouds are bad because they end in question marks.  Steve Krug wants us to design sites by eliminating the question marks from the users mind.

Some sources of question marks are:

  • Ambiguous or odd names
  • Is something clickable or not
  • Navigation issues

This principle can help us resolve one of the big issues in web design: How deep should my site be?  We have all heard the idea that "nothing important should be more than [insert arbitrary number] of clicks away from the home page.  In reality the number of clicks is far less important than the number of clicks with question marks.  When you have to think for a second before clicking on a link, that click is far more costly than the one where you click without thinking. 

One reason this idea is better than "X number of clicks away" is that a good number of your site users are not going to start on the home page anyway.  Clear navigation from anywhere, to anywhere is more important than how quickly a user can find his or her way from the home page to some arbitrary location on your site.

How do Real People use the web?

A friend of mine often says "people are strange... I am glad I am not one of THEM".  Often when doing web design we think about our users and try to imagine what they are like.  Since you are reading this on a web page (and I am typing it on a web page) by definition you and I are one of those strange creatures called "web users". 

So how do we use web sites?  Do we carefully read every word on every page? Do we carefully analyze the clever iconography that the graphic designers slaved over for hours?  Do we methodically visit each section of a web site in order: Introduction to Product, What can this product do for me, What is the best way to use this product, and finally Here's how to order?  Or do we do something else?  In reality here is how we use the web:

  • We don't read pages, we scan them
  • We satisfice- we don't weigh options but rather we take the first reasonable option offered to us
  • We don't figure out how things work, we muddle through

In other words, our audience uses our web site like a giant billboard.  Taking in the really big important stuff and ignoring the rest and making a snap decision about the relevance of what we see.  So Krug asks the question: If our users are going to use our website like a big billboard, what is a designer to do?  The answer is simple:  Build a really great billboard!

Here are some ideas to do just that:

  • Create a clear visual hierarchy on each page so things are easy to find
  • Take advantage of conventions
  • Break pages into clearly defined areas
  • Make it obvious what is clickable
  • Minimize the noise

Krug argues that much of the text on most websites is either Happy Talk or Instructions. The author has a rather brutal method for handling copy on web pages:

  • Get rid of half of the words on each page. Then get rid of half of what's left.
    • This reduces noise
    • Useful content becomes more prominent
    • Pages are shorter
Pet Peeve

Let me get a pet peeve off my chest.  At the top right of each page on this site is a disclaimer that these "book reviews" aren't so much book reviews as just my personal notes.  I think it's important to be upfront with readers about exactly what the author is providing.  So when Krug promises early in the book not to engage in site bashing but rather show us good examples of web sites, I believed him.  That is until about 1/3 of the way into the book when it struck me that nearly every example he was giving was of something that he didn't like on a web page. 

I personally think Krug has every right to criticize all kinds of web sites and designs, but he shouldn't do so after telling us that he isn't going to do that.  OK.. end of pet peeve! :)

Navigating

Navigation on websites is a field unto itself.  Any web designer will tell you that he or she agonizes over where to put the navigation and how to best convey navigation information as well as how to structure the site so navigation is as simple as possible.  Why is navigation such a difficult issue?  In a nutshell the internet is a medium that creates these unique circumstances:

  • No sense of scale (How big is a given site anyway?)
  • No sense of direction (Is this page left or right of the last one you were on?)
  • No sense of location (Where does this page site in the vast sea of web pages?)

Krug has an entire section devoted to navigation and it is highly recommended reading.  One of the nuggets he gives us is the analogy of Page Names to Street Signs.  It is important to know where you are at any given time, so give each page a name!

Random Thoughts
  • Trunk Test - If you were blind-folded, thrown in a trunk and let out on any random page of your site, how quickly could you find your way around?
  • Home page - Above all else, convey the big picture!
  • Don't ask for personal information until you have to.
  • The only answer for religious debates is user testing.
Testing

The last 20% of this book is devoted to the topic of usability testing for the web.  Krug takes the approach that cheap testing is possible and far more useful than no testing.  The only way to really know if your site is good is to do some usability testing.  Usability testing however is different than focus group testing. 

  • Focus groups - A group setting good for getting opinions and feelings about a product or site
  • Usability testing - Individuals that are observed doing tasks and gathering information on a site

Here is a list of things that Krug says he knows about testing:

  • Testing is an absolute must for a great site
  • Testing 1 user is 100% better than testing no users
  • Testing 1 user early on is better than testing 50 users late in the project
  • The importance of getting "representative users" is highly over-rated
  • The point of testing is not to prove anything but rather inform opinions
  • Testing is iterative
  • Nothing beats a live audience reaction
Conclusion

This is a great little book about web usability.  I loved the common sense approach along with the distilled principles that Krug applies.  Anyone doing web design would benefit from reading this book.  I strongly recommend getting a copy and putting it into your library.  (Incidentally, I read the copy that my boss had in his library but there is now a second edition out that you should probably get instead!)