EDITOR Fred Sampson [e-mail]

Or consider a major airline’s website that, even today, times out if the form goes untouched for a while. This is like a department-store clerk saying: “I’m sorry, you’ve been standing in the same area for too long. To continue shopping please leave the store and come back.” If you must start over, how likely are you to walk into the same store?

Business-wise, form design is not solely an issue of customer satisfaction and retention. Usability issues can affect the bottom line directly and extremely. The book describes a Web retailer whose revenue increased by $300 million after a small change to the checkout process (the company eliminated the requirement for shoppers to register before checking out).

A well-designed form is barely noticeable. But that doesn’t mean the design process is. Like most design problems, achieving a concise design that seems, in retrospect, obvious, requires much work. There are myriad factors to consider, each with multiple alternatives, each with its pros and cons. Should we have one long form or a sequence of short forms? Should help be provided in-line with the form controls, on the side, or on top? At what point should we validate fields and deal with errors? You could fill a book with the subtle-ties of form design.

Fortunately, Wroblewski has written that book. Drawing on years of experience designing for eBay and Yahoo, he has cataloged the major considerations involved in creating forms. He walks the difficult line between writing for novice and veteran designers.

For beginning interaction designers and de facto designers with nondesign job titles such as engineer, product manager, or founder, Wroblewski’s book serves as a primer on interaction design. It covers basics like thinking of the problem from the user’s perspective, rather than the machine’s. It covers the differences between GUI elements like the radio buttons and drop-down menus.

For veteran designers the book serves as a convenient reference and cheat sheet of things to consider. Before embarking on a new form design project it would be worthwhile flipping through the examples and advice to become re-sensitized to the issues. Plus, there are bound to be some insights and ideas that old timers hadn’t thought of.

I picked up a few nuggets along the way. One

of them pertains to the placement of labels next to fields. I already had a bias toward right-justified labels positioned just to the left of the fields, a practice Wroblewski affirms. What was news to me was how having labels above the fields is even more efficient. Citing an eye-tracking study, Wroblewski demonstrates that this arrangement allows the user to take in both the label and the field with a single eye fixation. It’s a good design choice for layouts that can afford the extra height this approach requires. It’s also handy when the form will be internationalized, because it leaves plenty of elbow room for the labels to stretch out when translated.

Another assumption Wroblewski challenges is that users must always register before they get to use our Web app. It’s a logical and standard way to proceed, but there are costs. We are making a demand from users before demonstrating value, the opposite of a normal sales process. It’s therefore a source of friction and a point at which we lose people. Wroblewski gives examples of a different strategy: Play first, ask questions later. Drop the new user right into the application to begin experimenting. No friction, no demands, no commitment required. If it works for them, they will be willing to register to save their data or to access other juicy features. Wroblewski calls this strategy “gradual engagement,” because we tempt the user with our product in stages.

Another discovery I found fascinating pertained to marketing questions like: “What’s your household income?” or “What is your date of birth?” These intrusive questions are on the form not for the user, but for the marketing department. For eBay, they were a major source of attrition, and removing them resulted in far more successful registrations. The fascinating part is that by asking the marketing questions later, after registration, significantly more people were willing to answer them.

As one would hope for a book written by a designer, it is highly usable. The tone is straightforward, clear, and friendly. It is loaded with example screen-shots, making it quick to read or skim. Every chapter ends by summarizing best practices.

The book is also thorough. When I began reviewing it, I made a list of design issues relating to forms to see if they were addressed. It fared well. In-line validation of fields as you go? Check. The perils of putting field labels within the fields themselves? Check. Strategies for revealing contingent fields? Check. Dynamically checking user names for availability and passwords for security as the user types?