Unless you’ve been on a technology fast the last few days, you’ve no doubt learned that working at Amazon is hell.

On Saturday, The New York Times published a widely shared article ‒ based on hundreds of employee interviews ‒ making the case that Amazon is using technology and modern management principles to push its white-collar employees past the breaking point. The result, as the piece conveys it, is a cutthroat, compassionless workplace on the verge of mutiny.

According to the article, 100-hour workweeks are the norm, the sight of crying co-workers is common, and work-life balance at the company is a joke.

Unsurprisingly, responses to the article’s revelations have been fast, loud and almost universally negative. Former employees have started sharing their own stories about the amount of work they were forced to contribute “no matter the emotional or physical toll.” Industry experts have argued that Amazon is just the canary in the coal mine and that other tech companies are just as bad if not worse.

Even the Onion chimed in with a satire of Amazon’s approach to solving its employee burnout problems.

The debate over the article has been thought-provoking and fascinating, but it has largely failed to answer what may be the most important question raised by the expose: Has data-driven management gone too far?

In this post, we will look at what Amazon’s drama can teach sales leadership about the fine line between data-driven workplace optimization and inhuman employee expectations, and answer the question that lies at the heart of this saga: How can modern businesses strike the right balance between using data to improve productivity and execution without dehumanizing and overworking the workforce?

Amazon’s Brand of Data-Driven Management

Almost since its inception, Amazon (and its founder and face, Jeff Bezos) has been considered cutting-edge for its use of data and technology in managing its workforce. According to the Times article, Bezos “turned to data-driven management very early” and has consistently run his company with an Ahab-like, unwavering “confidence in the power of metrics.”

How has this data-driven mindset manifested at Amazon? The chief way is in Bezos’ articles of faith, a collection of management philosophies designed to stave off mediocrity and hold employees accountable for the success of their projects.

“Within Amazon, ideal employees are often described as ‘athletes’ with endurance, speed (No. 8: ‘bias for action’), performance that can be measured and an ability to defy limits (No. 7: ‘think big’),” the article says.

While data-driven management, in a vacuum, can hardly be viewed as a negative, the recent revelations suggest that Amazon’s heartless pursuit of metrics is jeopardizing employees’ (and even the whole company’s) well-being. As the article from the Times makes clear, the pressure to get things right and push employees to achieve more and more could end up endangering Amazon’s ability to attract and retain employees.

“We always want to arrive at the right answer,” Tony Galbato, vice president for human resources, told the news source. “It would certainly be much easier to just compromise and not debate, but that may lead to the wrong decision.”

We’ve followed this unfolding story closely over the past few days and talked with a variety of sales leaders to understand what Amazon’s struggles can teach us about the right role for data-driven sales management.

Measuring Individual Employees

There is a pervasive view in the business world ‒ particularly in the individual results-oriented sales landscape ‒ that employee-specific metrics are most effective when they are used as a stick to punish underperforming employees.

This view is appealing ‒ transparency is a great way to inspire competition and encourage efficiency ‒ but it ultimately misses a crucial point about management: employee performance data is a great way to help employees reach their potential.

We thought about this principle a lot when we designed our sales coaching philosophy. Like most sales teams, we use activity metrics like dials made, meetings scheduled and deals closed to encourage healthy competition among our sales reps, but even more than that, we use it to identify areas where individual reps are weakest so that we can provide targeted coaching. Not to fire them, not to publicly embarrass them, but to help them focus on the areas that need the most attention so that they can improve their results.

This seems not to be the case at Amazon. According to Amy Michaels, a former Kindle marketer, “the company is running a continual performance improvement algorithm on its staff,” not, as the name would suggest, to teach employees how to be better at their jobs, but to encourage them “erode work-life boundaries, castigate themselves for shortcomings…and try to impress a company that can often feel like an insatiable taskmaster.”

This is the wrong way to use performance data. We have found that metrics-driven sales coaching works best when it is used as a learning mechanism for employees. If a rep is hitting her dials number but coming up short on her goal for scheduling meetings, for example, a review of her activity data can help her understand that it may be a good idea to focus on the conversations she’s having with prospects, not just the effort she’s putting into calling them. Maybe she can tighten her value prop, or perhaps she should focus on leaving more compelling voicemails.

The distinction between using employee data to spur them to work harder and using it to encourage them to work smarter may seem thin, but it is actually fundamental: Workers can learn to be better at their jobs, but they rarely do if they feel they are constantly on the verge of being embarrassed or fired.

“Disagree and Commit”

When faced with the allegations against his company’s hard-driving, contentious work environment, Bezos has had one main rebuttal: Harmony breeds groupthink and stifles creativity.

In fact, a few days after the initial story in the Times (for which he initially declined to comment), Bezos came out swinging against the accusations, saying that he doesn’t “recognize this Amazon and I very much hope you don’t, either.” Instead, he argued that he has always wanted to create a work environment that used disagreement and plainspeak to get the most out of employees, not beat them down.

This distinction is a crucial part of the story, and, again, one that seems to be overlooked in the commentary around the article. There is no denying that unpopular opinions (especially when backed by data) are one of the most common sources of true innovations. The question, though, is how these unpopular opinions are broached, and who voices them and how.

Again, this is an incredibly interesting argument to have in the context of sales. Historically, sales has been a field governed by the gut. Sales managers and VPs often hold beliefs for years (always be closing, control the conversation) that may not be rooted in data.

The advent of business intelligence and democratized sales data have, thankfully, helped get to the bottom of many of these issues. And in the process, they have stirred the pot a bit, and put some sales leaders in the hotseat.

This type of disagreement is good. Business leaders should always follow the data, even when it flies in the face of their long-held beliefs, but this is quite different than the way Amazon has been “managing by the numbers,” (at least according to the recent press).

It is easy for managers and executives to hide behind the screen of “accurate but impolite” when they are forcing employees to respond to emails 24 hours a day or openly criticize a co-worker. The truth, though, is that this is a misrepresentation of what it means to encourage “diverse viewpoints.”

Too often, it appears that Amazon’s management was relying on this interpretation to create a one-way street of constructive criticism, which is, of course, the opposite of diverse viewpoints. If data is going to be used to challenge the status quo and hold workers to a different standard, it must be used in both directions. Sales reps, not just executives, must feel empowered to voice their opinions and make data-backed cases for different ways of doing things.

The Future of Data-Driven Management

Few people deny that business management will become increasingly data-driven as time wears on. The influx of easy-to-use business intelligence and improved technology make this essentially inevitably.

The news about Amazon, however, has complicated this evolution for some. Fears about over-reaching managers and the continued disintegration of work-life balance were brought to the fore by the revelations about Amazon’s culture. But they should not be viewed as the death knell of data-driven sales management.

Like any tool, data-driven management can be used for good or bad. It is up to you and your company to make sure you’re not simply relying on performance metrics to beat employees into submission, but as a way to provide targeted coaching and to encourage a higher level of discourse at an organizational level.