Using test/control methods is primarily about understanding the way of thinking rather than knowing any specific comparison framework or statistical test. It’s often an art figuring out which factors to isolate. Ideally, the test is based on some hypothesis about the underlying behavioral dynamics. Especially in B2B, we don’t always get a giant sample to hold out, so being specific and focused about what to test is critical.
This chef gets it. His hypothesis: ripening tomatoes lose moisture around the stem. A little tape and a few days proves his point, so store them upside down. How did I grow up in a town with a Tomato Festival Parade and still not know this?
JP Snow 