If you’ve spent any time around engineers and software developers, then you’ve surely heard many forms of the same a question — a question that is also common in hierarchy-dominated companies without agile techniques:

Does management ever do anything?

Does it seem like there are more managers than employees?

Why do we need managers at all?

On the other hand, if you work at an agile business, you’ve probably grown accustomed to working in networks with self-employed, self-motivated contractors. You’re probably comfortable following leads, directing projects, and dealing with people who use their own resources to handle things. Agile workplaces are definitely different from the old-school hierarchy, but the same kind of question still pops up:

Do we even need a manager?

Is there a skill managers have that justifies authority?

Who died and made that guy boss?

In Short: Does Management Work?

100 Years In The Making

Since the 1911 publication of Frederick Winslow Taylor’s The Principles of Scientific Management, businesses have been formally trying to determine and follow what they call “best practices.” But what are today’s best practices, especially now that modern business is more about innovation than assembly lines?

A recent study by Nicholas Bloom of Stanford University proves that, if you’ve ever asked what management does, you may have good reason to wonder. Bloom’s research group, including Raffaella Sadun of Harvard Business School and John Van Reenen at the London School of Economics, tested whether thousands of businesses followed practices considered to be essential to good management.

After compiling 10 years’ worth of studies from 100 researchers in 20 different countries covering 8,000 medium-sized manufacturing firms with between 50 and 5000 employees, the results were unequivocal: In Bloom’s own words: “Poor management is rampant.” That’s probably not a surprise to anyone, except the managers themselves.

The 3 Essential Elements of Good Management

Bloom found these three best practices were essential to good management:

1. Targets

Does the organization in question support long-term goals with tough—but achievable—short-term benchmarks?

2. Incentives

Does the organization reward high performers? Do they retrain or move under-performers?

3. Monitoring

Does the organization collect and identify performance data to identify opportunities for improvement?

Bloom used those three elements and ranked the 8,000 companies in the study on a scale from 1 to 5—5 being the best—where each point represented a 23% rise in productivity, 14% increase in market capitalization, and 1.4% annual sales growth. Scores of 5 were rare; scores of 1 were plentiful.

The Prognosis Is Not Complicated – Nor does it need to be

Unfortunately, a staggering majority of businesses don’t know how to develop targets, create incentives to reach those targets, and monitor and measure progress.

As Bloom states, “A call for ‘better management’ may not seem like a cutting-edge idea, but given the potentially large effects on incomes, productivity and delivery of critically needed services worldwide, it may actually be a radical one.”

I think agile practices—with open communication; a combination of scrums, sprints and data visualization, and emphasis on emergent strategies with a “try anything once” mentality—are the answer.

Awareness is the First Step

In the same way that awareness is the first step in customer purchase, awareness is also the first step in improving bad business management. Business managers either don’t know they have a problem, or they can’t be objective in evaluating their own skill level. It seems bad managers spend too much time lobbying for how good of a manager they think they are.

For example, 79% of the 8,000 organizations in the study claimed to have “above average” management practices. That’s 79% of cross-section of overwhelmingly poor business managers, whose self-assessment of management ability had no correlation to their measured score in Bloom’s study. No correlation!

Not only does this expose management’s typically rosy opinion of itself, it also suggests the human tendency—like the children of Lake Woebegone—to believe, “We’re all above average.”

“Above average.” It’s not the best tagline, considering how low the bar sits to begin with.

Criticisms of Bloom’s Study

Bloom’s research, which he also discussed on Freakonomics Radio, episode “I Consult, Therefore I Am” , showed management can improve. In fact, improvements come fairly easy. When companies in the study received five months of heavy business consulting versus only one month of light consulting, the heavy consulting produced dramatically better business results. But many of the companies Bloom studied worked in textile and steel mill environments in places like India, where simple improvements like “inventory control”—having some, rather than none—and “equipment maintenance” produced vast improvements over the control group.

Agile marketers more likely work in different environments, like technology or building SaaS software. What good could specific management initiatives like those do for us?

Not much, on the surface.

And what about business consultants?

If you don’t have a high opinion of business management, you probably don’t have a high opinion of business management consultants. Freakonomics host Steven J. Dubner lists the objections to consultants:

  • Many consultants tell clients what they want to hear.
  • Clients hire consultants to justify decisions they already want to make.
  • Clients hire consultants to have a scapegoat in case those decisions don’t work.
  • Business consultants have no qualifications or certifications.
  • Business consultants have little actual experience in business strategy.

By one funny (if sexist) summary: a management consultant is “a guy that can tell you 50 ways to make love but doesn’t know any women.”

Spare Us The Celebrity Figurehead

In a second Freakonomics podcast, “How Much Does A good Boss Really Matter?”, Dubner references studies that show a good boss can produce upwards of 10% gains in worker productivity. That’s the boss’s contribution alone, and 10% is substantial. But these are direct bosses. Hands on. A coach who played the game. On the other hand, figurehead leaders—like celebrity CEOs, presidents, and other bigwigs—don’t matter as much. Their contributions to worker productivity according to the study measured much lower.

Good Bosses Matter Most in Agile

Whether agile or not, good management works. Good management is rare, but the way to be good is rather simple:

  • target goals
  • create incentives
  • monitor progress

And if you accept the history that shows lean manufacturing, cost cutting measures, efficiency, and widespread deregulation have all reached a point of diminishing return, then agile appears to be the only way forward. It’s up to technology industries, and specifically SaaS—arguably the most competitive industry on the planet—to lead the way.

  • Our process is the leanest
  • Projects come and go faster
  • Markets change faster
  • Fortunes are made and lost faster
  • People jump on and off bandwagons with more regularity
  • Agile software development is where agile management ideas came from in the first place

But how can we create the networks required to be agile, especially when many of us face entrenched (and measurably poor) management hierarchies? The answer: A dual-operating system, where agile networks establish themselves over time alongside the hierarchy.

Checklist: Do You Follow The Three Essential Elements of Good Management?

These questions were adapted from interviews in the study of over 8,000 manufacturers in 20 countries. For more questions from the study, see worldmanagementsurvey.org


  • How do you communicate organizational goals to individual workers?
  • Does anyone complain that the target definitions are too complex?
  • How do you deal with repeated failures in a specific business segment?


  • How do senior managers show that attracting, developing, and retaining talent is a top priority?
  • How long is underperformance tolerated?
  • What makes it distinctive or special to work at your company?
  • What does your company do about star performers who want to leave?


  • How do problems get exposed and fixed?
  • What key indicators do you use for performance tracking?
  • For a given problem, how do you identify the root cause?