''How do you test the name of a product, such as a cola drink or a software development methodology?''

One way is to ask people if they like the name. You could even ask them to come up with a better name. Some will like the name, some won't. You'll get a list of approximately 1.65 proposed names per person you ask.

A better way is to present the product to different (statistically significant) populations of subjects, keeping all variables constant except the name.

If 55 percent prefer your cola under the name of Barfo, and 89 percent prefer it when the name is Ralpho, the name Ralpho may be presumed to be better than Barfo.

----

''When should you test the name of a product, such as a cola drink or a software development methodology?''

When the effect of the name on sale of the product is stronger than any other change you could imagine making to the product.