The Average Customer Doesn't Exist

In January 2002, Kmart filed for Chapter 11 bankruptcy protection. What makes the story worth retelling is what happened just before. In the months leading up to the filing, Kmart was posting record customer-satisfaction scores alongside $32 billion in sales.

Dambisa Moyo tells this story in her book How Boards Work, and her diagnosis is precise. The high scores came mostly from Kmart's lower-spending customers. Hidden inside the average was a growing disaffection among the higher-spending customers the company depended on for revenue. Kmart had no way to see their satisfaction separately, so neither management nor the board knew. As Moyo puts it, more disaggregated data would likely have revealed the problem, and might even have helped avert bankruptcy.

I have read a lot of corporate case studies over thirty years. Few capture the cost of standardization as cleanly as this one.

An average is a decision about what to ignore

Every average is a choice. When you blend thousands of people into one number, you decide in advance that their differences don't matter. Sometimes that's a reasonable shortcut. Often it's the very thing that blinds you.

Kmart's leaders weren't careless. They were measuring what most organizations measure, the way most organizations measure it. The system told them customers were happy. The customers who mattered most were quietly walking out the door, and the system had no way to say so.

This is why I keep coming back to the individual. Organizations are built to see segments, demographics and averages. People don't live as segments. They live as individuals, with their own reasons for staying and their own reasons for leaving. When your measures can't see the individual, you aren't really seeing your customers. You're seeing a statistical portrait that none of them would recognize.

The same blind spot lives inside your walls

It would be easy to read Kmart as a customer story. I think it's a leadership story, and it applies to your people as much as your market.

Most engagement surveys report a single score. Most culture dashboards report averages by department. A team can look healthy on paper while its most capable people are deciding to leave. The average holds steady right up until the moment it doesn't, and by then the people you most needed to keep are gone.

Our own research through the Workplace Serendipity Assessment shows how much an overall picture can hide. People are still executing well. Underneath that, we see them struggling to grow and to share, because environments don't feel safe enough to connect the dots or trust their own judgment. A single score would tell you everything is fine. The pattern underneath tells you something else entirely.

Three questions for any leader reading the numbers

Who is inside this average? Before you celebrate a score, ask whose experience it actually reflects. If the answer is "mostly the people who are easiest to satisfy," you don't have good news yet.

Whose silence would hurt most? Every organization has people whose departure would change its future: a critical customer, a rising leader, a quiet expert. Find out whether your measures can hear them specifically.

What would we do differently if we knew each person? That's the real question behind personalization. Not "how do we segment better," but "what changes when we stop treating people as interchangeable?"

Seeing people one at a time

Moyo's broader point is that boards make decisions with imperfect information, and that the quality of the information shapes the quality of the decision. I agree, and I would add one thing. The most dangerous gaps aren't missing numbers. They're numbers that look complete while flattening the individuals behind them.

Standardized measurement built a great deal of what works in modern business. I'm not arguing to throw it out. I'm arguing for the healthy tension between the standard and the person: keep the measures that let you run the organization, and add the ones that let you see the people you're running it for.

Kmart had all the data it thought it needed. What it lacked was a way to see the customers who mattered most as individuals.

This essay responds to ideas in How Boards Work by Dambisa Moyo (Basic Books, 2021). The Kmart example and its analysis are hers.

Want to explore these ideas further? Learn more about my work on leadership, identity and conviction at www.theglennllopis.com.