Thomas Edison "failed" 10,000 times before he perfected the light bulb. But the inventor considered each failed effort a success. He had eliminated 10,000 combinations of materials that didn't work.

Edison treated his endeavors as experiments. If he had viewed the results as finished goods, he probably would have given up after a few tries, concluding that incandescent lighting was just a bad idea.

Today's businesses can profit by Edison's example. When companies skip the experimentation phase, they in effect use the marketplace or their own organizations as a lab. Promising new products and innovative business practices are launched before they're proven seaworthy. No wonder two-thirds of reengineering efforts fail and fewer than 900 of the 16,000 products introduced in 1993 are still on the shelves.

People fail to experiment.

Would you buy an expensive new suit without making sure it fit? Companies make a much costlier mistake when they fail to experiment. Instead of beta testing the idea in some small part of the organization or market place, they do a full blown implementation. An experiment is a trial run. You do a small test of the idea to determine if your hypotheses (i.e. assumptions) are valid.

In fact, if you're doing it right, you should be trying to prove your hypotheses are false, that your assumptions are INVALID! You should be trying to figure out why reengineering isn't for you, why customers won't want your new portable widget.

People fail to get valid feedback.

The whole reason for doing an experiment is to get information about your assumptions and hypotheses. Only accurate, relevant, objective feedback will produce useful information. Remember you're running an experiment ... unless you're incredibly lucky, the feedback is going to tell you the experiment FAILED.

None of us likes to hear we've failed. In fact, many of us refuse to hear it, rearrange the facts, or discredit the source: "The idea is great! What do the customers know?"

The next time you test a product idea, ask your prospects, "What's the worst thing about this product?" "Why is this the worst product idea you've ever seen?" You need tough love, not honey-coated feedback. If you don't hear the truth at the beginning, the marketplace will send you the message later, and it will cost you more than a temporary blow to your ego.

People don't learn from their experiments.

Assuming you're in the experimenting mode and have gotten good feedback, you now have to use that information to prepare for your next experiment ... yes, another experiment. Using the information you gathered, you must adjust your hypotheses and modify your ideas. Eliminate some alternatives, try others, and keep experimenting until you get it right. Remember how long it took Edison. One experiment doesn't prove an idea or justify a product.

Edison had a lot of success with failure. He was wise enough to record and use the data his experiments produced. That's how a bright idea becomes a brilliant achievement.