Customer Learning vs Machine Learning

Have you ever wondered why that google search on baby clothes you just finished resulted in a new Facebook ad featuring the latest line of baby clothes from Buy Buy Baby? Or how about that time you were singing the words to Don't Worry Be Happy and all of sudden there was a suggestion to buy the "Best of Bobby McFerrin" in your Amazon cache. Coincidences? No Way!

Ever since the days that "Customer Loyalty Cards" were invented within the grocery store chain businesses have companies been clamoring for a way to get inside the consumer's head (err . . . their buying habits and choices). Anticipation of the customer's predicted behavior is an incredibly powerful tool. And, one that comes with incredible outcomes and consequences.

Sure. We all want first class service with every shopping experience. Who wouldn't value a cart of all their normal food selections waiting for them when they visit their grocery store's website? Think of the time savings, the hassle, the personalization, the ease of the shopping experience.

Now, how about walking into your favorite restaurant and not having to order a thing . . . . the chef and waiters already know via an alert when you walk in the door about your preferences, allergies, and if it's a special occasion or not. Too far? Too personable? Buckle up . . .

Machine Learning is the concept of giving a computer the ability to build algorithms that will anticipate a predictable outcome. Like: a consumer buys diapers + baby food + car seat = suggestion to buy a crib. This is the evolution of the consumer experience; like it or not. Eventually, the buying process will get faster, more automated, and seamless in terms of transactions.

Now, what needs to be emphasized is the Customer Learning side of the equation? If a consumer knows that their buying behavior will be tracked and calculated by a machine which will ultimately generate a suggestable selling scenario, shouldn't we be smart enough to dictate the outcome???? Or at least have some sort of input into how the data is gathered?

Or, are we ok with the simplicity/decision-relief in acquiescing our purchasing to the output of a computer program?