What Paper is (and Isn't) Good For - Paper Prototyping: The Fast and Easy Way to Define and Refine User Interfaces - page 209

What Paper Prototypes May Find (Look Issues)
When it comes to look, you first need to examine the questions you have that are specific to the visual
aspects of the interface. As described in the previous section , for many types of questions, the look
plays only a secondary role. But naturally many aspects of the visual design can be important
depending on the circumstances. Here are three examples.
Of Interest ... Look and Brand—Not Inseparable
Look isn't always the most important factor in conveying a brand. I once tested two versions of a
prerelease financial Web site. The first version tested had all the navigation and most of the text but
lacked the graphic design—all pages had black text on a white background with no images. In other
words, it was functional but ugly. The second version, which was ready a couple of months later, added
the professionally designed colors, fonts, and images of smiling people.
We tested each version with half a dozen users in each of two populations (consumers and small
business owners). At the end of the session, users were given a list of adjectives and asked to agree
or disagree with whether they described the site. The adjectives represented the brand that the
company wanted their site to have: reliable, accurate, honest, and so on. Interestingly, the brand
perception didn't seem to change very much after the graphic design was added—users perceived the
same flaws in the site's content and organization despite the fact that the second version looked much
nicer than the first.
(Caveat: This study had a small number of users and there were other differences between the two
versions besides the graphic design. In other words, this is an anecdote, not scientific proof. But it's
intriguing that we saw no evidence of what we expected—that the brand perception would be improved
by the addition of professional graphic design. I would love to see more formal research to help us
understand when and how graphic design has the sort of positive effect that we instinctively believe it
has.)
Alignment. One paper prototype used a tree structure with multiple levels of indenting. Because
the items in the hand-drawn paper prototype didn't line up very well, users had to work harder to
understand the hierarchy that was being represented. After the test, one of the users commented
that it probably would have been clearer on a computer screen, and I agreed with him.
Font size. A colleague told me a funny story about font size in a Web site intended for use by
teenagers. After their initial launch, the designers discovered that the large font they'd chosen
could be read by a parent halfway across the room, which was unacceptable to the teenage target
audience!
Photographs. Some interfaces have photographs as part of the content, such as a picture of a
sweater on a clothing Web site. If photographs are important, you can paste them onto an
otherwise hand-drawn interface. (This is what The MathWorks did with their prototype described in
Chapter 2 .) On the other hand, if the interface uses photographs for a primarily decorative purpose,
it's probably safe to omit them from a paper prototype.
Other things that are hard to simulate to a high degree of accuracy with a hand-drawn prototype include
colors, shading, contrast, readability, white space, icons, dense layouts, and complex graphics.
Because I find that this is pretty self-evident to most people, I won't belabor the obvious by providing
lots of examples. If you have important questions that depend on specific visual elements like these,
then a hand-drawn paper prototype probably isn't going to answer them to your satisfaction. (And if
there are enough differences between how your design appears on paper as compared with on the
screen, maybe printed screen shots aren't good enough either; in that case, it would be important to
test the design on a computer.)