The hamburger menu experiments by James Foster are interesting, but the problem is that they assume the primary goal is navigation clicks. Rarely is a navigation click a bottom line business metric for a company. Almost always you want the user to convert into a customer for whatever you are selling. A more compelling experiment would be one that shows results for both a primary action of new visitor to customer as well as nav click conversions. The fact that fewer users click on the navigation is not necessarily a bad thing.