The Customer Resume

Allow me to tell you a brief story. Let's go back to November 1990. Yes, 1990. I was in my first week of work at Lands' End. One of the aspects of my job (statistical analyst ... today that job would be titled a digital analyst or a data scientist or a growth hacker) was to audit my statistical models.

I submitted a request each Friday to print the customer purchase resume of one of every ten-thousand customers. Over the weekend, the customer file was scored, and when I arrived at work on Monday, a stack of printed customer records magically appeared on my chair. Every Monday morning, I thumbed through each record, hand-calculating that my scores were being applied properly.

The best part of the process was to watch how customer behavior evolved. Not overall customer behavior, but rather, the behavior of an individual customer. I can still remember customer # 1000004 ... her record was always on the top of the stack (no, I am not going to share her name with you, though I can still remember her name). She hadn't purchased in a long time, so every Monday there was drama ... had she reactivated over the past weekend?

Every marketing employee should be required to look at a customer resume. Every one.

What do you see when you look at the resume at the beginning of the post? Be honest. Study it. How does this customer behave? This isn't much of a "digital" customer, is it? Or a seasonal customer. But this customer is a very, very good customer, make no mistake about it. So enjoy this customer - quit demanding that this customer engage with you digitally. Collect profit. Ok?

By the way, there was a Friday back in 1992. I submitted my request to print one out of every 10,000 customers ... except I typed in 1 in every 100 customers. On Monday morning, our IT team, who clearly saw the request and knew I made a mistake, elected to initiate a "teachable moment" and dutifully printed the resume of one out of every 100 customers. They deposited the 40,000 sheets of paper in my office. You learn a lot when you audit your model scores, don't you?