While I agree with Krug’s assertion that users of mass market e-commerce sites will vary widely in their likes, dislikes, motivations and preferred navigation styles, this does not negate the need for careful selection of usability test participants. Krug suggests "grabbing some people" at random for testing, but this approach may backfire if the participants do not share the navigation style or goals of the critical user bases for the site under development. While his suggested approach to testing and recruitment may be cost effective for testing navigation structure or familiar transactions such as a shopping cart application and is certainly better than performing no testing at all, it may have expensive consequences for the design of complex, high risk web applications – especially within the environment of a corporate intranet.

Here’s an example: let’s say you were designing a web based application for a high-use call center that required users to look up complex information while speaking to your customers on the phone. Your call center is located in the Midwest, but your developers work in your corporate headquarters in Northern California. Since the developers have no call center experience, and little understanding of the relative importance of various parts of the application, and since they consider the Midwest call center personnel to be relatively unskilled computer users, they decide to create an easy to learn mouse-driven interface requiring several clicks to move between screens. Using Krug’s approach to usability testing, the developers go down the hall and recruit a few web-savvy colleagues to “test” the interface. Sure enough, the designers and developers participating in the test are able to complete particular tasks, and the usability test is declared successful. But when the application is launched, the call center experiences a slowdown in workflow because critical screens are difficult to reach, and the increased reliance on using a mouse for navigation results in an increase in repetitive stress injuries. Both issues result in increased costs to your company, and furthermore, your customer satisfaction ratings plummet as a result of the slowdowns in the phone queue!

When participants in a usability test are not representative of the end users of a web driven application, important aspects of its usability, such as the context and frequency of use, are likely to be overlooked. Additionally, the view of the users held by the developers of a system may be flawed, and those misperceptions can have disastrous results for the usability of your web site. Despite Krug’s assertion that “it doesn’t matter who you test”, our experience has shown that including representative users in the testing of high impact, high cost web applications provides an important reality check in the development process.

Despite these shortcomings, the readability of Don’t Make Me Think! makes it a valuable book, especially for those new to the idea of designing usable web sites or who need convincing that usability testing is an important part of the design process. If you are having difficulty getting the decision makers or developers in your organization to understand why usability is important, I recommend giving them a copy of this book and suggesting they read it on their next plane trip....

I enjoyed it very much and I think I would send one to a couple of very experienced web designers who just don't "get" that websites are designed for real people to be used in everyday life and not to be some sort of experimental work of art that noone has access to. Great job, Mr. Krug!

Others have labeled this book basic, "for beginners"; it is a simple message, but think about the sites you frequent, doesn't the title sum up your experiences there?

In much the same way Roy Williams gives the number crunchers of advertising a reality check in "The Wizard of Ads", Steve Krug does with web site usability. The book is fluid easy reading. It provides suggestions for usability testing on a budget, and quick, simple, usability ideas for those who may have become bogged down in the latest technology and bloated design.

Large-scale site development requires solid information architecture, layout, and design. This book is no substitute for that process, but it masterfully delivers on what it claims to be - "A Common Sense Approach to Web Usability"