In one of today’s lectures, we briefly discussed association analysis - the relatively simple idea behind grocery store loyalty cards and product suggestions like Amazon’s “customers who bought A also purchased B, C and/or D”. The idea is to look at a customer’s entire order and use this information to increase sales through strategic suggestions. The discussion sparked a 9-year-old memory of a philosophy class I took in college, particularly the section on Kant and ethics/morality.

Immanuel Kant developed a set of decision rules for determining whether or not an act was moral. His first rule, the Formula of Universal Law, is the one that interests me today. It basically requires you to look at an act from a perspective of cold logic and ask yourself: “if everyone else did this, would it be possible?” If the answer is no, the act is immoral.

So what does this have to do with product suggestions?

Let’s imagine that I buy a pen on Amazon, but I don’t know what type of notebook to buy to go with it. I decide that I will purchase whichever notebook is listed first on the product suggestion list. Was that a moral decision?

According to Kant’s Universal Law, no. If every other customer in the market for both pens and notebooks chose to use the same decision method, there would never be any notebooks suggested.

Of course, this is an extreme and silly example. Customers do make their own purchasing decisions and so suggestions are nearly always available. But just know that every time you follow the advice of the Amazon suggestions, Immanuel Kant judges you a little.