Do, Redo & Undo

When creating something, it's easier to think in the affirmative. We think in a vector of taking actions and building things, and can forget that over time undoing those same decisions can be just as important. Do, Redo & Undo asks a group to focus on this, and to think through the implications of dismantling and altering.

This is a useful exercise in developing any human-to-machine or human-to-human system. Software provides myriad cases of undoing: users need to change configurations, fix mistakes, and remove software entirely. Business processes need to address this equally well: components need to change or dissolve, and often this flexibility is lost without clarity on how it is done.

Small groups

1 hour or more, depending on the complexity of the existing "Best-Case Scenario"

Generally, the group would run this exercise after they have a concept or prototype as a starting point. In the case of software, it may be a user story or feature list; in a process, it may be a draft of the flow.

The group should be given time to walk through and digest this example. The exercise opens with the group brainstorming answers to a simple question: "What mistakes can and will be made?"

Using Post-Up, the group brainstorms a set of items on sticky notes and pools them to create a starting set of scenarios to explore "undoing and redoing." It's not unusual for a few humorous items to make the list. Other questions to consider asking in fleshing out the set include:

In generating the initial list in Post-Up, the group has identified at least one Worst-Case Scenario. Their task now is to address the items by focusing on three possible solutions:

A group that has a large number of items in the Worst-Case Scenario may wish to prioritize them by likelihood and then focus on the hotspots. There is an implied order of preference in Do, Redo & Undo. A problem that can be entirely eliminated by changing the design avoids needing a "redo" or "undo" solution. For example, a feature that asks the user to enter her contact information might be eliminated entirely, if the information can be fetched from somewhere else.

As the group works through Do, Redo & Undo, they should capture their solutions and revisit the original Best-Case Scenario. Their draft of solutions should accompany the design as it matures, eventually proving itself in user testing and the real world.

The Do, Redo & Undo game is credited to James Macanufo.