Note: We created two separate designs for user testing, one which took a more user-friendly approach and another which focused exclusively on efficiency. After the first round of testing (which consisted of three music directors and an "elf" from the radio station) we dropped the design which focused on efficiency and merged the best design concepts from both into a single design. Because we ran each user in the first round on both designs (we counterbalanced so that an equal number of users started with either design) there was a significant learning and preference bias (the tasks remained the same across the two designs, therefore users were more comfortable by the time they tried the second design). This may have been an influence on the users' unanimous preference for the second design they were presented with. Nonetheless, we found it a useful exercise to use multiple designs in our first round of testing, since many details from both made it into the final design.
...