Got an impossible problem? Need fresh ideas? Looking for that breakthrough concept? Brainstorming may be the solution to all these things and more.

If you've had bad luck with brainstorming, you're just not doing it right. Because in my experience, as long as you have the right group of people, holding a brainstorming session is like throwing a match into a room full of firecrackers. There's a sudden and powerful chain reaction.

Here are a few suggestions for creating some fireworks of your own.

Before Your Session

Select a leader. You need someone who will remain realistic and low-key, stifle negative statements, and keep the group on track.

Define your problem. Just write a sentence or two that clearly defines the subject your group will address.

Create an agenda. List the topics you want to cover and any specific brainstorming techniques you want to try. Prepare a few ideas in advance to serve as a starter and be prepared to suggest questions to help spark ideas.

Set time limits. It's usually best to keep it short, about 15 to 45 minutes. If you go longer, take breaks every hour to keep people fresh.

Set quotas. The idea is to work fast and produce lots of ideas that will be evaluated at another time. You can set an overall quota or individual quotas for each topic. Coming up with 100 or more ideas isn't hard. Just two ideas a minute will add up to 120 in an hour.

Select your participants. A mixed, energetic group of the same general rank is best. Don't invite authority figures who may stifle the group's openness.

Circulate background information. Announce your session (never call it a "meeting") and prime the participants with your problem statement, background information, and examples of the kind of ideas you're looking for.

During Your Session

Review the problem. Don't put people to sleep; just quickly go over the problem, background data, and what you hope to accomplish. If there are questions, answer them before you begin.

Establish the ground rules. Decide what can and can't be done and read your rules to the group. For example: Participants WILL contribute lots of ideas, share every idea that comes to mind, and encourage out-of-the-box thinking. Participants WILL NOT criticize any idea, hold back an idea because it may sound silly, or evaluate ideas during the session.

Take detailed notes. You need someone who can write fast. An option is to record the session and transcribe notes later.

After Your Session

Allow time for more ideas. If you've had a productive session, ideas will continue to occur to people for hours or days afterward. Ask everyone to write down these ideas and submit them.

Type up all ideas. The final product of a session will be a multi-page document that lists every single idea created. Nothing should be edited. Organize or classify these ideas in some fashion for later evaluation.

Evaluate and choose. It's often best for those responsible for the problem to evaluate the ideas, though you can run into "idea ownership" problems. On the other hand, another group may not be able to grasp the significance of many of the ideas generated. You'll have to experiment. Either way, I suggest a smaller evaluation group - fewer than six. Don't rule out crazy ideas too quickly. Allow yourselves to develop the ideas as needed.

When the dust settles, you'll probably find yourself with a bunch of really stupid ideas, but so what? You'll also have a handful of surprisingly good ideas. And the whole process can energize your staff and improve morale.

Don't get discouraged if it doesn't work perfectly. It often takes time to assemble the right group, create an open atmosphere, and produce the best results. As with so many other things in life, practice makes perfect.