Amazon, Data, and Selling Your Culture Down the River

Over the weekend I was sifting through a survey from Navex Global about compliance training programs, and came across what chief compliance officers consider the biggest obstacles to effective training programs: employee cynicism about corporate culture, and fear of speaking up.

Then I turned to the New York Times and read its brutal, brilliant, portrait of corporate culture at Amazon.com—a company that somehow took simple, noble principles of good corporate organization and perverted them into something awful, into a business that looks like a miserable place to work. A rank-and-yank system of performance reviews that pits one employee against another. An employee feedback system that generates anonymous personnel tension far more than it roots out legitimate misconduct. A culture so obsessed with data that executives turn a blind eye to context, which breeds ever more fear of failure into an overworked and exhausted workforce.

And employee cynicism and fear are the biggest obstacles to effective compliance and a strong corporate culture, you say? Gee, I can’t imagine why.

This profile shows everything that is wrong with senior executives who glom onto buzzwords like “data driven management” and “performance metrics” without considering how those ideas can trample over human nature. Amazon has created an environment that sounds fantastic when it’s nothing more than a list of values and principles on a web page somewhere, and then fails all the hallmarks of a culture that really drives trust and change for the better.

Of course the corporate spokesmen Amazon have mounted a counter-offensive, saying the company looks nothing like what the New York Times depicts. Nevertheless, the article (and the thousands of reader comments it has generated as well) should be required reading for ethics & compliance officers everywhere. Then summon your fellow senior leaders to ask: are we in danger of becoming like this?

The problem is best summed up in this statement from one current Amazon employee, defending Amazon’s exhaustive tracking of data and applying those metrics to personnel performance. “Data creates a lot of clarity around decision making,” the man is quoted as saying. “Data is incredibly liberating.”

Well, no it’s not, unless you mean that data lets the company abdicate responsibility for making a difficult choice. Having information at your fingertips can help inform the choices you make, certainly; that’s wise management. A slavish devotion to data, however—which seems to be what Amazon exalts above all—removes the need to choose. It reduces all questions to binary calculations; the very nature of data is to value something numerically as 0 or 1. And when the environment is binary, everything that sounds like a yes-no question to management feels like a win-lose proposition to employees.

That’s when you lose trust. That’s when you breed cynicism, because employees know delivering even 0.99 results will still be marked as “not 1,” which therefore means 0. That’s data-driven management taken to its logical, dehumanized conclusion.

It gets worse. Amazon also tries to embrace a speak-up culture, with online tools that let employees provide anonymous feedback on peers and meetings where executives are required to debate and find flaws in others’ ideas. The tools themselves are only tools, harmless unless someone uses them maliciously. Rigorous debate at meetings is also harmless—it’s praiseworthy, actually—so long as people on the losing side of the debate aren’t punished for trying.

But people on the losing side are punished when a data-obsessed culture boils away all factors and only see the failure. Anonymous reporting tools are used as weapons when data determines your fate. It’s a rich irony: Amazon’s tools to deliver a better culture become the vehicles employees use to create a worse culture.

It’s a rich irony: Amazon’s tools to deliver a better culture become the vehicles employees use to create a worse culture.

What’s even more disappointing is that the flaws in Amazon’s approach aren’t difficult to see. Every statistics or social science worker knows the Fundamental Attribution Error: our tendency to put more emphasis on internal characteristics for behavior (this person failed because he’s lazy) rather than external ones (he failed because the tools didn’t work, his wife died, he was sick, and so forth). Some of the most popular business books available (look it up on Amazon if you must) come from the brothers Chip and Dan Heath, who wrote Switch: How to Change Things When Change is Hard about nurturing change at large organizations. One of the most important factors, they say, is that employees must know they can try to change and fail, and will not be punished.

All of what Amazon does, if the New York Times article is accurate, flies in the face of that wisdom.

In the spirit of Amazon’s focus on data, let me end with one final piece of it. I am connected on LinkedIn to one person currently employed there, and to six who left. With a culture like anything like that, I can see why.

Comment on this post on LinkedIn.