Use trails for information graphics: Replacing designers' assumptions with feedback from users

This paper illustrates lessons we learned from user trials. Our brief was to design product graphics to encourage fast and accurate use of a new point-of-sale laser scanner.Our early intuitions were proved wrong. We had thought that the most effective design would explain how the scanner works. But our tests showed that this confused users. So instead we concentrated on guiding people to successful behaviour, without concern for the way the scanner works.Tests with typical users indicated the best graphic - simple, familiar, conspicuous, and easy to name. It significantly increased the acceptability of the product and improved the speed and accuracy with which people used it, even though it encouraged an incorrect model of the way the scanner works. This finding has implications for the design of graphics and metaphors for all kinds of user interfaces. Additionally, user trials proved invaluable in challenging our mistaken assumptions and in guiding us to an effective solution.