Turnover Never Appears on the P&L. That Is Why It Costs So Much.

The most expensive people risk on your P&L is the one your accounting system was never designed to show you.

A departure never arrives as an invoice. It costs what it costs across four different budgets, four different months, and almost never against the name of the person who left (SHRM; MyCulture). By the time anyone adds it up, the number belongs to last year — and to nobody.

The four lines hiding inside one departure

• Recruitment cost. Fees, advertising, screening hours, interview time — the most visible part, and the smallest one.
• Onboarding time. The weeks before the replacement produces what the predecessor produced — measured in their salary, paid at full rate.
• Lost output. The work that did not happen while a seat was empty, or half-full, and the overtime that absorbed it.
• Manager hours. The most expensive line nobody tracks: exit conversations, recruitment rounds, training, and the team's meetings that turn into therapy sessions.

The details differ by source; the order of magnitude does not. The Romanian comparators are built from local payroll and replacement data (Zarina CRM; BusinessIMM), and they land in the same band, which is what turns “50–200%” from a slogan into a planning assumption.

Your number, in thirty seconds

The calculator's default turnover, 20% a year, is the Romanian private-sector benchmark cited further down; the €18,000 average salary is a placeholder. Change both to your own; the arithmetic is the one your CFO will use when asked.

The result is a range, because replacement cost varies with role, market and timing (European benchmarks; AmCham Romania). A range with your own headcount in it still beats both opinions: the CEO who says turnover is normal, and the HR lead who says it is expensive.

Why it stays invisible

The mechanics are mundane. Finance tracks what arrives as a transaction: an invoice, a payroll line, a fee. Turnover arrives as an absence — productivity that did not happen, hours that went elsewhere, a team that slowed down. Absences do not generate documents. So the cost is paid in full and recorded nowhere (cost breakdowns for Romanian employers include exactly these categories among the hidden ones).

The pressure behind it shows in the 2026 data: in Randstad's 2026 HR Trends study, 28% of employees who resign cite excessive workload as the trigger — a cause no salary line fixes and no finance report records (Randstad Romania).

Four questions to ask before the next budget round

1. What did our five most recent departures actually cost? (If the answer takes a week, that is the answer.)
2. What share of last year's exits happened before month 18 — the window where selection, more than the market, tends to be the cause?
3. Which managers lost more than one person in the last year? That is where the next resignation is already priced in.
4. What would a 20% reduction in turnover be worth on our own calculator numbers? Compare it to the cost of the program that would achieve it.

The bottom line

Turnover is an operations cost with an HR label: paid by the P&L and recorded by no one. The companies that put a number on it get a defensible figure, and a defensible figure is what lets a company stop arguing about whether retention deserves budget and start deciding how much of the number to remove.

Our Organizational Diagnosis puts a price on the turnover your accounts do not show and traces it to where it starts: selection and the manager layer.