How to measure turnover properly
An overall company figure is almost useless: it averages healthy and unhealthy areas together. You need to break it down.
Turnover, % = (number of leavers in the period ÷ average headcount for the period) × 100
Then break that figure down:
- By department. Often 70% of all departures come from a single team — and that is a question for a specific manager, not for the labour market
- By length of service. Count separately those who left in the first 3 months. That diagnoses hiring and onboarding, not retention
- By type of departure. Resignation or employer-initiated dismissal are fundamentally different stories
- By value of the employee. A weak employee leaving is healthy. A strong one leaving is a loss. In the overall figure they look identical
What level counts as normal
| Sector | Rough annual norm | When to intervene |
|---|---|---|
| Retail, food service, high-volume roles | 30–50% | above 60% |
| Manufacturing, logistics | 15–25% | above 35% |
| Office, specialists | 10–15% | above 20% |
| Managers | 5–10% | above 15% |
| Departures in the first 3 months | up to 10–15% | above 20% |
The real reasons people leave
In exit interviews almost everyone names salary — it is the safest answer and blames nobody. But dig deeper and the picture looks different.
The immediate manager
The most common real cause. People join a company and leave a boss. If turnover in one department is three times higher than in a neighbouring one under identical conditions, the issue is almost never money.
The gap between what was promised and what is real
Someone was sold one job and received another: development was promised and routine delivered; a team was promised and they turned out to be doing everything alone. Such departures concentrate in the first three months and relate directly to how the vacancy was described at interview.
No visible future
The employee does not understand what will happen to them in a year: what levels exist, what earns promotion, what they need to be able to do. With no answer, the choice is simple — look for prospects elsewhere.
Overload after someone else leaves
A classic chain reaction: one person resigns, their work is divided among those remaining, and two months later a second one leaves. This is exactly why a long-open vacancy is dangerous for more than idle time.
Money
A real factor, but more often a trigger than a root cause. If someone finds the work interesting and comfortable, an offer with a 10% increase rarely outweighs that. If they are already dissatisfied, it outweighs it easily.
What genuinely reduces turnover
Measures in order of return on effort
- Deal with your managers. Look at turnover by department. If the problem is localised, work with that specific person rather than redesigning the whole company's incentive scheme
- Describe the job honestly at interview. Including the unpleasant parts. A candidate who knows about the difficulties in advance does not run from them a month later
- Work on the first 90 days. This is where the largest share of departures concentrates, and it is the most manageable area
- Introduce a monthly one-to-one. Thirty minutes on what is getting in the way and what comes next. Most departures can be predicted a month or six weeks ahead if these conversations happen
- Describe career steps. Even a simple scheme of "what you need to be able to do for the next level" retains people better than an unplanned pay rise
- Do not have exit interviews run by the manager people are leaving. Nobody tells the truth to the person who caused their departure
How to predict a departure in advance
A decision to leave does not ripen in a day. Usually one to three months pass between the internal "that is it, I am leaving" and the resignation letter — and during that period behaviour changes noticeably.
Early signs
- Stopped offering ideas or arguing — does exactly what is assigned and no more
- Reduced involvement in general discussions and the informal life of the team
- Started strictly observing working hours after a period of overtime
- Frequent time off and mid-day absences appear
- Stopped talking about plans beyond the coming month
- Updated their professional profile or became noticeably more active in industry communities
One sign means nothing. Three or more at once is grounds for a personal conversation — not framed as "I hear you are planning to leave" but as "how do you feel about the work, what is getting in the way, what is missing".
In a company of 60 people, 14 employees resigned over a year. Overall turnover of around 23% looked almost normal for the sector, and nobody raised the issue.
A breakdown by department showed something different: 9 of the 14 departures came from a single team of 12 people. In that department turnover was 75%, while across the rest of the company it was under 10%.
The cause turned out to be neither salaries nor the market but the management style of the department head. After working with him and changing how tasks were set, departures in that team fell to two people the following year.
Exit interviews: how to get the truth
Most exit interviews are useless because they are conducted formally and by the wrong person. A few rules that change the quality of answers:
- Not conducted by the manager the person is leaving. Otherwise you will never get an honest answer
- Hold the conversation after all documents are signed, when the person has nothing left to lose and need not worry about references
- Questions about facts, not opinions. Not "did you enjoy working here?" but "at what point did you first think about leaving, and what had happened then?"
- Keep the promise of confidentiality. If what was said is repeated verbatim to the manager, future employees will say nothing
- Accumulate the results. One interview is an individual case; ten interviews over a year is a picture of the company's problems
What turnover costs
A company of 50 people with 25% turnover loses and replaces 12–13 people a year. With a conservative replacement estimate of 3–4 monthly salaries and an average salary of 350,000 ₸, that is around 15 million tenge a year — money that appears nowhere as a separate budget line.
Reducing turnover from 25% to 15% in such a company saves roughly 6 million tenge annually. That makes the payback on onboarding, management training and quality selection fairly obvious.
Where to start this week
- Calculate last year's turnover for the company and separately by department
- Identify separately those who left in the first three months
- Find the department leading on departures and address that one specifically
- Draw up a list of 5–7 employees whose loss would hurt, and speak to each of them personally
In brief
Average turnover in the country is such that roughly one in five employees leaves the company, and for 48% of employers this is the main task of 2026. The metric must be broken down — by department, length of service and value of the people. The main causes lie not in salary but in the manager, mismatched expectations and absence of prospects. The greatest return comes from working with managers and building proper onboarding for the first 90 days.