PHASE 1: THE NEXT MAJOR RELEASE

By the time the usability lab and team were created, the decision had been made that the next release of Lotus cc:Mail was to be a major release, not a point release. The plan was to produce an entirely new and improved version of cc:Mail, including many new features. From a usability perspective, the ideal goal for a product with such a significant installed base would have been to retain the good aspects of the UI and improve the bad. Instead, the development plan was to throw it all out and start over fresh. Since the underlying functionality was to be significantly different, this fact was to be reflected in a brand new cc:Mail UI.

A committee called UA/UI (User Agent/User Interface) was in existence at this point in the process. The User Agent referred to the "client", or the code that presented the UI of cc:Mail. The members of the UA/UI committee included development representatives from the five cc:Mail platform teams ( Windows, Mac, OS/2, DOS, and OPEN LOOK), managers, technical writers, product management, and the one usability engineer. Approximately nine people attended a typical UA/UI meeting. One person was designated to take minutes, but no other roles were established. For several weeks, the committee met eight hours a day, four or five days a week, gradually reducing to half- or full-day meetings once or twice a week.

The UA/UI committee was planning the next major release to be the first version to run on Windows 95 (still called V4 or Chicago then). This new environment meant designing for SDI (Single Document Interface), instead of MDI (Multiple Document Interface). It also meant a new focus on "objects" (OLE2). Since the UA/UI meetings focused on these technologies, they were heavily system-centered, as opposed to task-oriented or user-centered.

A typical UA/UI discussion revolved around types of message lists and storage methods. Whenever a visual or behavioral UI design issue was raised, it was postponed, as the developers could rarely agree on a cross-platform UI solution. For example, if a graphical representation proposed for Windows did not have an obvious DOS alternative, or had a conflicting meaning in OS/2, the discussion was noted in the minutes and put on hold indefinitely.

UA/UI meeting minutes, mainly questions, open issues, and investigative action item assignments, were typed and distributed, read, revised, and redistributed. Additional UI design and development discussions took place via e-mail and in hallway discussions. The modal of the UA/UI meetings was clearly "design by committee". After a while, the usability engineer was asked to run these meetings. Due to this intensive time commitment, usability testing was not yet taking place in the new lab.

Beginning of Document | Top of This Page (Phase 1)
Phase 2 | Phase 3 | Phase 4
Results and Examples | Conclusions