This book is a quick
read. The author says he worked for
that, looking for something that could be covered in a long plane ride.
I bought it in
This book is expensive.
$35. ($24.95 if you go here,
and the author gets a commission!)
This book is valuable.
I’m working on projects right now where I will put its lessons,
especially the “testing on the cheap” outline in chapters 9 and 10 to work
Mr. Krug’s experience
is directed toward web sites where users “do something,” although, as he
says, the insights can also be applied to those that are mostly about publishing
information. (And, yes, I will apply
the principles to this site, real soon now …!)
The fundamental principle, as expressed in the title, is to make the
purpose of the site and the steps to accomplishing important tasks completely
obvious. Failing that, at least make
The author suggests that
most of us see web sites like bill boards
at 60 mph: quick glimpses of the
highlights as we zoom by. On the
other hand, those who develop web sites often view them as literature or works
of art: to be read with some care.
Therein lies a problem. The
graphic on p. 23 of “What designers build …” and “What users see…”
captures the essence wonderfully. The
designer’s screen is sharp and clear. The
user’s side has two screen shots, each with just three elements in clear
focus: the name of the site on both,
then two elements determined by what the user’s purpose is in visiting the
a/k/a/ “good enough” is the order of the day for most of us in most things.
We find a way to do something and, if it is good enough, we don’t
invest time in looking for improvements. As
with scanning instead of reading, this makes for less than careful study of the
options and text on a web site.
Almost as a sub-point of
satisficing, we muddle through rather
than figuring out how things work. Does
anyone really use all the settings on their dishwasher, washing machine, or any
So, what’s the best
response from a web design team?
a clear visual hierarchy:
conventions (unless you’ve got a good reason not to)
“home page” is a special problem because everyone wants a piece.
Do the best you can with the compromises.
Test early, test often.
It’s an iterative process of design, prototype, test, analyze, design,
prototype, test ….
Three users or less will
do, and need not be too carefully selected from your “target” group (unless,
of course, that’s easy to do). Even
when you can focus on target group, go outside it.
You can never be sure ALL members of group will COMPLETELY share a COMMON
Start testing in the
planning stage by testing competitor’s sites.
Then test rough sketches, page designs, prototypes, the first usable
version, and run new pages by the person in the next cubicle, at least.
A computer (or the paper
design), the user, a facilitator, and a videocamera, with a monitor in a nearby
room for other team members to watch. See
for a script for a testing session and a sample release form.
Again, chapters 9 and 10
are worth the price of the book if you’ve never been exposed to this kind of