PProve
Home Guides › People decisions by the numbers
The research

People decisions by the numbers.

Hiring, promoting, and holding on to the right people are the highest-consequence calls a leader makes — and most are made on a gut feel. Here’s what the research actually says about how often those bets go wrong, and what it costs when they do.
By James CarterUpdated July 17, 2026
70%

of the variance in a team’s engagement traces to a single factor: its manager. (Gallup)

20–213%

of annual salary to replace one employee — ~20% for most roles, up to 213% for senior or specialized ones. (Center for American Progress)

75%

of voluntary turnover is preventable — most people leave for reasons an employer could have seen. (Work Institute)

11%

of executives strongly agree their leadership-development programs achieve lasting results. (McKinsey)

Every figure here was checked against its original publisher before publication (see References below). This page is Be Legendary’s synthesis of publicly available research — not a proprietary survey. Statistics belong to their respective publishers.

1. The person you promote is the lever — for better or worse.

Gallup’s analysis of millions of teams found that roughly 70% of the variance in a team’s engagement is attributable to one factor: its manager. Put plainly, if you know nothing about a group except who leads it, you can predict a surprising amount about how it performs. That makes the decision of who gets to lead the single most leveraged people call you make — and the one most often made on the thinnest evidence.

The usual evidence is past performance. But strong performance in one role is a notoriously weak predictor of success in a different one — the Peter Principle in action. The behaviors that made someone a great individual contributor aren’t the behaviors a larger, more ambiguous role demands. Knowing who to promote means reading behavior the current role never tested.

2. When the bet is wrong, the bill is bigger than it looks.

The Center for American Progress, synthesizing decades of studies, estimated the typical cost of replacing an employee at about 20% of their annual salary — and as much as 213% for senior leaders and hard-to-replace specialists. That’s lost productivity while the seat is empty, the cost of hiring and onboarding, and the slow ramp before a replacement is fully effective. For a lean company, one wrong bet on a key role can erase a quarter of progress.

Most of that cost is avoidable. The Work Institute, drawing on more than 100,000 exit interviews, has found that as much as 75% of voluntary turnover is preventable — people leave for reasons an employer could have seen and addressed, most often a lack of visible growth. The departures that hurt most are the ones you’d have taken back: regretted attrition.

3. Spending more on development isn’t fixing it.

U.S. companies pour roughly $14 billion a year into leadership development. Yet McKinsey found that only 11% of executives strongly agree those programs achieve and sustain the results they’re meant to. The problem usually isn’t the curriculum. It’s aiming it blind — investing in everyone equally, with no way to tell who is actually compounding and who is coasting, and no signal for who was worth the bet in the first place.

Read together, these numbers point to one thing. They aren’t a people problem — they’re a measurement problem. Every one of them is what happens when a high-stakes bet gets placed on the wrong evidence.

4. What the evidence points to: measure the behavior first.

Performance tells you what someone has already done. Engagement tells you how they feel. Neither tells you what they’ll do when the work gets hard, ambiguous, or unglamorous — which is exactly what a bigger role, or a retained key player, is a bet on. That’s commitment behavior, and unlike a survey score it’s observable: initiative when no one’s watching, follow-through under pressure, and learning from being wrong.

Measuring that — over time, in real conditions — before you promote, hire, or hand over the keys is the one change that moves a people decision from a guess to a bet you can defend. It’s what the Commitment Quotient is built to do.

References

  1. Gallup. State of the American Manager (analysis of 2.5M+ work units). Managers account for ~70% of the variance in team engagement. gallup.com
  2. Boushey, H. & Glynn, S.J. (2012). There Are Significant Business Costs to Replacing Employees. Center for American Progress. Typical turnover cost ~20% of salary; up to 213% for executive and highly specialized roles. americanprogress.org
  3. Work Institute. Retention Report (100,000+ exit interviews). As much as ~75% of voluntary turnover is preventable. workinstitute.com
  4. Gurdjian, P., Halbeisen, T. & Lane, K. (2014). Why leadership-development programs fail. McKinsey & Company. U.S. companies spend ~$14B/year on leadership development. mckinsey.com
  5. McKinsey & Company (2016). What’s missing in leadership development? Only 11% of 500+ executives polled strongly agreed their leadership-development programs achieve and sustain the desired results. mckinsey.com

Third-party statistics are the property of their respective publishers and are cited here for commentary. Figures reflect the most recent editions available at publication and will be updated as new data is released.

Stop betting on the wrong evidence.

Take the free 2-minute Certainty Diagnostic to see how much you’re guessing — or book a call to see how Prove measures commitment behavior where it counts.

Keep reading