Optimal Brainstorming refers to a set of heuristics for productive brainstorming sessions informed by creativity research as well as the experience of the CIA in directing analyst groups in scenario and idea generation. It subverts a variety of biases that get in the way of producing good results.
The key idea is the finding that quantity trumps quality when it comes to idea generation. It was found in multiple trials that groups that were encouraged to produce more total ideas always beat groups attempting to produce high quality ideas in producing what was later judged to be the highest quality ideas by blinded groups of judges. This is especially relevant since idea generation is so cheap. All we really care about is getting a single hit, regardless of the number of misses.
We therefore enhance all aspects that improve quantity while removing any roadblocks.
The most important guidelines:
The exact way of meeting these guidelines can vary, but probably looks something like this:
A bounded space and time, ideally about 90 minutes, in a space with whiteboards, stickynotes, or other affordances for visualizing or rearranging ideas on the fly. Negative capability is enhanced by holding off on proposing solutions in favor of exploring the problem fully. All particiapants know that the current session is explicitly not for solving the problem. Psychological safety is enhanced by encouraging curiosity, humor, and playfullness. The initial stage of the brainstorm is done individually rather than as a group to avoid the pitfall of only gathering ideas from the loudest speakers, or having the activity immediately derailed.
The facilitator gives a prompt or prompts related to the problem and has the participants each write their ideas on a pad of paper. Prompts should include a sentence or two about turning off the internal censor, explaining the quantity over quality idea. The facilitator should not halt the individual stage at the first sign that everyone is out of things to write, this is often exactly the stage at which original thinking is about to happen after the obvious ideas have been exhausted.
Everyone shares some number of their best ideas, ideally on sticky notes so they can be clustered in the next stage.
Once all the ideas are up, everyone takes a few moments to soak in everyone elses' ideas. Suggestions for categories are then solicited by the facilitator, and different clusterings and labels of the ideas are tried, to see if there is any obvious larger scale structure to the ideas generated. Relations between categories might suggest themselves.
Open discussion should follow. The facilitator should take steps to encourage contribution by more quiet participants, as eveness of communication is a predictor of group success. This process may solve the problem outright, or generate new prompts for further sessions. When deeper analysis is desired, the results of an initial broad overview may be split up and assigned to several smaller teams of 3-5 people to search more deeply in parallel.