Most frontline managers get the job because they were the best at the work. Then the work changes. Before the promotion, their output was tickets closed, first response time, renewal rate or quota. After it, their output is whatever the team produces. That’s a different job, and hardly anyone gets trained for it.
The numbers on that are blunt. In CMI and YouGov’s 2023 research in the UK, 82% of people moving into management had no formal management or leadership training when they started. In West Monroe’s 2018 survey of 500 US managers, 59% of those overseeing one or two people had received no managerial training at all. And in a SHRM survey of US workers, 84% said poorly trained people managers create a lot of unnecessary work and stress.
None of that arrives on the P&L with a label on it. It arrives in four places.
1. Escalations that climb
When a frontline manager isn’t sure they’re allowed to make a call, the call climbs. The credit over the usual limit, the SLA exception, the angry customer, the rep who wants a schedule change. Each one lands on a director or on you, which means the most expensive time in the company goes to decisions a trained manager would have closed in a minute.
That time is already scarce. Only 30% of leaders say they have enough time to carry out their responsibilities with the depth and diligence required (DDI Global Leadership Forecast 2025). And the managers in the middle are stretched in the wrong direction: middle managers spend nearly half their time on non-managerial work, including almost a full day a week on admin (McKinsey, 2023).
An untrained manager never shows up as a line item. That’s exactly why the cost keeps growing.
2. Turnover you pay for twice
People leave managers. In BambooHR’s 2025 survey, 58% of people who left a job in the last year said their boss’s management style was the main reason, up from 37% in 2017.
You pay for that twice: once for the person who leaves, and again for the months it takes their replacement to get back to where they were. The recruiting bill is the small part. SHRM’s benchmark puts the average cost per hire at $4,683 (SHRM, 2022), and that covers recruiting costs only. Research cited by the SHRM Foundation puts the total cost of turnover, once lost productivity and ramp time are counted, at 90% to 200% of annual salary (SHRM Foundation, 2008).
3. Work done twice
Rework is what happens when nobody wrote down what good looks like. A rep closes a ticket that comes back. An implementation hands off to success without the context the customer already gave twice. A new hire does the job the way the last person showed them, which isn’t the way the next person expects.
The root is usually the standard, not the person. Only 10% of US workers say job descriptions always accurately reflect the reality of the role (Monster, 2026), and 26% of new hires say they would restart their job search over mismatched expectations (ZipRecruiter, 2026). Role ambiguity is linked to lower job performance (Tubré and Collins, Journal of Management, 2000). A manager who can’t say what good looks like can’t measure it, coach it or delegate it.
4. Performance lost while you wait
Every month a new manager runs untrained has a price, and the clearest measurement of it comes from one company’s own evaluation. Verizon ran a two-day program for more than 2,000 first-time managers and compared trained managers against a matched untrained group. The trained managers improved 2.1% more across five business metrics, the program returned 415% annualized, and delaying the training by a year gave up an estimated $8,552 in performance per manager (Paul Leone, Training Industry Magazine, 2019).
That’s one company, measured by its own ROI consultant, so treat it as a strong example rather than a law. The broader evidence points the same way. A meta-analysis of 335 studies found leadership training increased on-the-job leadership behavior by 28% (Lacerenza et al., Journal of Applied Psychology, 2017).
A note on the numbers. Every figure in this piece comes from a named source, linked where it appears. Survey figures describe the people surveyed, and some samples are global or UK rather than US. The 28% is the paper’s own percentile-based conversion of its effect sizes, not a raw productivity gain.
Put a number on yours
You don’t need a consultant to see the size of this. You need an afternoon and four questions.
- Escalations. How many hours a week do your senior people spend on decisions a frontline manager should make? Multiply by their loaded hourly cost. As a purely hypothetical example, two directors each losing three hours a week at $100 an hour is $600 a week, or about $31,000 a year.
- Turnover. How many people left in the last year because of their manager? The exit interview usually says so. Multiply by your cost to replace one, ramp time included.
- Rework. How many tickets or tasks were reopened or redone last month? Multiply by the average time to redo one.
- Waiting. How long does a request sit waiting on a manager’s decision in your highest-volume queue, and what does that delay cost the customer on the other end?
Add them up. That’s the operating cost the problem is creating. One honesty rule: count cost avoided and time recovered, and nothing else. Don’t count revenue you might earn with the freed-up hours, because nobody can honestly promise it.
What actually lowers it
Not a one-day workshop on its own. The same meta-analysis found that practice, feedback and sessions spaced over time were each linked to stronger results, which is why training that sticks looks less like an event and more like a few weeks of work on the manager’s real team.
The pieces that move the four cost lines are specific:
- A written standard for every role, so rework has a definition and performance has a baseline. That’s the role doc.
- Clear decision rights, so escalations stop at the right level. That’s the manager charter: what you own, what you decide, what you escalate.
- Management and leadership, named separately, so a manager knows which half they’re avoiding. That’s Keep The SCORE, Earn The TRUST.
- AI on the admin, never on people decisions. That day a week of admin is exactly the work an AI tool can compress, as long as it stays an administrative staff officer: company-approved tools, placeholder names for employees, and anything touching performance, pay or separation through HR.
- Measurement against each manager’s own baseline, so you can see what changed, not just whether people enjoyed the session.
Untrained managers aren’t a people problem you solve with a pep talk. They’re an operating cost, and like any operating cost, you can measure it, lower it and check that it stayed down.