A resignation letter almost never names the real reason. The exit survey says better pay, shorter commute, bigger title. Then, over the next year, three more people from the same team hand in the same letter, and the org chart tells the story the survey did not. The common denominator is not the company, the industry, or the compensation band. It is one desk: the direct manager’s.
The research on this is unusually consistent. DDI found that 57 percent of employees have left a job because of their manager, and another 32 percent have seriously considered it. Gallup’s turnover research found that 42 percent of people who quit say their manager or organization could have done something to keep them, and that 45 percent of departing employees had no proactive conversation about their satisfaction or future in their final three months on the job. The door was open the whole time. Nobody walked through it.
This is not a Great Resignation leftover. The Bureau of Labor Statistics counted 3.2 million quits in June 2026 alone, a 2.0 percent monthly rate. Even in a cooler labor market, voluntary turnover is a multimillion-person event every single month, and the leavers skew toward the people you least want to lose. Replacing a leader costs around 200 percent of salary by Gallup’s estimate, a technical professional around 80 percent, a frontline employee around 40 percent.
Yet when executives budget for retention, the money flows to the transactional levers: compensation adjustments, retention bonuses, perks. McKinsey documented this mismatch during the biggest quit wave in modern history. Employers assumed people left for pay. The leavers themselves ranked relational reasons first: not feeling valued by their organization (54 percent), not feeling valued by their manager (52 percent), and no sense of belonging (51 percent). Pew found the same pattern from a different angle: 57 percent of 2021 quitters cited feeling disrespected at work, nearly matching the 63 percent who cited pay. Respect is not delivered by a policy. It is delivered, or withheld, by a manager, one interaction at a time.
Gallup’s landmark finding, drawn from data on 27 million employees, is that the manager accounts for at least 70 percent of the variance in team engagement. No other single factor comes close. Companies will commission a culture survey, redesign the benefits package, and rebrand the careers page before they will ask the harder question: which of our managers are quietly driving people out?
The problem is getting worse, not better, because the managers themselves are struggling. Gallup’s 2026 global report shows employee engagement fell to 20 percent in 2025, the second straight annual decline, and managers posted the sharpest engagement drop of any employee group. Gallup put the cost of declining manager engagement at 438 billion dollars in lost productivity in 2024. Disengaged managers do not develop people, do not hold career conversations, and do not notice a flight risk until the calendar invite says “quick chat.”
Here is the stat that should stop every executive team: only 44 percent of managers worldwide have ever received formal management training. Most organizations hand someone eight direct reports, a performance review deadline, and zero preparation. Employees notice. In SHRM’s research, 84 percent of American workers said poorly trained managers create unnecessary work and stress, 57 percent said their own manager needs people management training, and half said their personal performance would improve with a better trained boss. The top gaps workers named were the basics: communicating effectively (41 percent) and developing the team (38 percent).
The upside runs just as steep. Gallup found managers trained in coaching skills show up to 22 percent higher engagement themselves, and their teams show up to 18 percent higher engagement. Few investments in the entire HR budget have a measured effect that large.
There is a specific version of this problem that life sciences, manufacturing, defense, and technical services all share: promoting the best individual contributor into management because they were the best individual contributor.
The evidence here is blunt. Researchers Alan Benson, Danielle Li, and Kelly Shue studied promotions across 214 firms and found companies consistently promoted their top performers, yet “the better the salesperson, the worse they were at managing.” The skills that made someone exceptional at the bench, the line, or the account are not the skills that make a team want to stay. Technical organizations run this play constantly. The best scientist becomes lab head. The best machine operator becomes shift supervisor. The best cleared engineer becomes program lead.
Then the organization compounds the error by skipping the training, and often by skipping the conversation entirely. DDI found 70 percent of frontline leaders were not even expecting their promotion when it came, and 18 percent regret taking the role. So the typical technical manager is someone who did not plan to manage, was never taught to manage, and is still measured mostly on technical output. Predictable behaviors follow: they keep doing the old job and treat people questions as interruptions, they hoard decisions because they trust their own judgment most, they give feedback the way a reviewer marks up data rather than the way a coach builds a person, and they never ask anyone where they want their career to go.
The team reads all of it as indifference, and indifference is a quit driver. The organization ends up losing twice. It lost its best contributor’s output the day of the promotion, and it loses the team over the following two years.
Life sciences. U.S. life sciences employment hit a record 2.1 million in early 2025 by CBRE’s count, concentrated in a handful of clusters like Boston, the Bay Area, and the Washington to Baltimore corridor. That geography matters: when a bench scientist decides their lab head does not develop people, a competitor is hiring within commuting distance. Lab leadership is also where the promotion trap runs hottest, because academic training produces brilliant scientists who have never once been taught to run a one-on-one. When a discovery team loses two senior scientists in a quarter, program timelines slip in ways no signing bonus repairs.
Manufacturing. The Manufacturing Institute’s retention study with the American Psychological Association found workers stay mainly because they enjoy the work (83 percent) and value stability (79 percent), and that career development is the top concern for workers under 25 (69 percent). All three of those run directly through the plant supervisor, who controls assignments, schedules, and who gets tapped for the next certification. The same study found that involving frontline workers in decisions closed about two thirds of the satisfaction gap between the floor and the front office. That involvement is a supervisor behavior, not a corporate program.
Defense. In NDIA’s Vital Signs 2025 report, 54 percent of government respondents and 41 percent of industry respondents ranked workforce issues among the most pressing challenges facing the defense industrial base, and the report notes clearance requirements shrink the talent pool while raising the cost of every hire. That math changes the stakes of manager quality. When a cleared systems engineer quits over a program lead who never gives feedback, the replacement is not a 60-day search. It can be a year of recruiting, adjudication, and onboarding while the program absorbs the gap. On cleared teams, one weak manager is a schedule risk, not just an HR issue.
Service industries. In high-volume service operations, the frontline supervisor is the company for most employees. Gallup pegs frontline replacement at around 40 percent of salary, which sounds manageable until you multiply it across a contact center or field operation running 30 percent annual turnover, much of it clustered under specific supervisors. Pew’s finding that disrespect drove 57 percent of quits lands hardest here, because scheduling, break coverage, and basic dignity are all daily supervisor decisions.
Manager capability feels fuzzy until you instrument it. Three practices turn it into data.
Cut engagement scores by manager, not just by department. Most companies aggregate survey results at the division level, which hides exactly the signal that matters. A division at the 60th percentile can contain teams at the 90th and teams at the 15th, and the difference is almost always the person running each team. Trend it over time and manager problems surface a year before the resignations do.
Run structured skip-levels. A skip-level is a conversation between a leader and the people two levels down, without the middle manager present. Done casually, it produces pleasantries. Done with a consistent question set (What would make you leave? When did your manager last discuss your career with you? What does your manager not know about this team?), it produces early warning. Remember Gallup’s 45 percent figure: nearly half of leavers had no meaningful conversation in their final three months. Skip-levels are how leadership finds out those conversations are not happening.
Read exit data at the cohort level. One departure is a story; a pattern is a diagnosis. Track regrettable attrition per manager over a trailing 24 months, alongside first-year attrition of their new hires and internal transfer requests out of their team. Individual exit interviews are polite and vague. Cohort data is neither.
They build a real dual career ladder. The only durable fix for the promotion trap is a technical track with the pay, prestige, and titles to match management, so a principal scientist or master machinist can advance without inheriting direct reports. If management is the only way up, you will keep converting your best contributors into your weakest managers.
They select managers on manager criteria. Gallup estimates about 1 in 10 people have high talent to manage, and that companies choose the wrong person 82 percent of the time, usually by rewarding tenure or technical output. Sophisticated organizations screen candidates for the actual job: developing people, communicating, making decisions under ambiguity, and wanting the role in the first place.
They train before day one, not after year three. Since only 44 percent of managers ever get formal training, providing it is a competitive advantage on its own. The measured payoff of coaching-skills training, up to 22 percent higher manager engagement and 18 percent higher team engagement, is the return case in a single sentence.
They make retention a managed number per manager. What gets reviewed gets attention. Put regrettable attrition, engagement trend, and internal mobility on each manager’s scorecard, review it quarterly, and treat a bad trend the way you would treat a quality escape: with a corrective plan, not a shrug.
They require stay conversations. A twice-yearly structured conversation about what keeps each person, what would pull them away, and what they want next converts the 42 percent of preventable exits from a statistic into a to-do list. The conversation costs an hour. The alternative costs 80 to 200 percent of a salary.
First, demand attrition and engagement data cut by manager. If your CHRO cannot show you regrettable turnover by leader for the trailing two years, your organization is managing its biggest retention variable on anecdote.
Second, fund the dual ladder and fix manager selection. Every technical promotion decision is really two decisions: who gains a manager and who loses a contributor. Make both on purpose.
Third, treat manager training as risk management, not development spending. At 200 percent of salary per lost leader and 80 percent per lost technical professional, a modest capability program pays for itself the first time it prevents one team from unraveling.
The companies that win the next decade of talent competition will not be the ones with the flashiest benefits. They will be the ones where a strong scientist, operator, or engineer looks at their manager and sees a reason to stay. That is buildable. Most of your competitors are not building it, and that is the opportunity.
RX2 Solutions is a workforce solutions firm specializing in HR outsourcing, executive search, and strategic staffing. We partner with organizations to build high-performing teams through customized talent strategies, leadership placement, and scalable workforce solutions.
📞 Phone: 610.340.3490
📧 Email: info@rx2solutions.com
🌐 Website: www.rx2solutions.com
RX2 Solutions
A Respectfully Professional People Company