"You know, we recently had a soft drink product (an exotic berry seltzer line) go through one of those simulated test markets, and it was a disaster. The new products people forgot completely about the possibility that the customers who bought the product in the shopping center pseudo stores might not actually get around to trying it. But it happened. Based on in store purchases, everything was OK, but a good percentage of the purchasers changed their minds later; and, if they used the product at all, it was limited trial by just one person. Solution, of course: A sales wave test added to the end of the store test. But that increases the cost considerably¬. Could you tell me when we should use the added sales wave and when we shouldn't?"