Why workers leave in the first 90 days

Ask a plant HR head where their attrition concentrates and the answer is always the same: the first quarter. The operator who completes ninety days usually completes the season. The one who leaves almost always leaves before day ninety. Most companies respond with the only lever on their desk, a wage revision, and are surprised when the leaving continues. The wage was never the problem. The arithmetic of settling was.

Day one: the deposit and the broker

A worker who travels a thousand kilometres to your corridor arrives with a fixed sum and a fixed goal, a number he has promised to send home each month. Before he has worked a single shift, the local housing market takes its cut. A room in an unmanaged PG wants a deposit, often two months of rent, and the broker who found it wants his fee. The capital the worker arrived with, the buffer that was supposed to carry him to his first wage, is gone in the first week. He starts your employment already borrowing.

Day thirty: the leak

The first wage lands and the worker does the sum he came to do: earnings, minus room, minus food, minus the electricity charged at whatever rate the landlord invented, minus the commute, minus the remittance commission. What reaches home is a fraction of what he promised. He does not tell your supervisor this. He tells the man from his village on the next cot, and both of them quietly begin recalculating whether the migration pays.

Day forty-five: the body keeps score

An operator living ninety minutes from your gate spends three hours a day commuting on top of a full shift. He sleeps six to a room, eats what is cheap rather than what is food, and drinks water nobody has tested. The result shows up on your floor as the things you already measure: late punches, unplanned absences, slower cycle times, defects in the last hour of the shift. Fatigue is a housing outcome long before it is a discipline issue.

Day ninety: the maths fails

By the end of the first quarter the worker has run the experiment. If the savings are not accumulating, the rational move is to go home, and he makes it. Your plant records another regretted exit, opens another requisition, pays another round of sourcing and training, and runs the line short while the seat is empty. Then the replacement arrives, pays a deposit to the same landlord and a fee to the same broker, and the cycle begins again. This is why NITI Aayog's S.A.F.E. Accommodation report treats worker housing as manufacturing infrastructure: the country is short roughly 25 million worker beds, and every one of those missing beds is being paid for somewhere as attrition.

What actually breaks the cycle

Wage increases move the top line of the worker's arithmetic. Settlement support moves everything below it, and everything below it is where the migration fails. A worker who lands into managed accommodation near the plant skips the deposit, skips the broker, walks to work, eats three predictable meals, and pays one bundled price with nothing invented at month-end. His savings start in the first wage cycle instead of month four. That is the entire retention mechanism: workers stay where their savings grow. At Nia, roughly 7 in 10 members are still with us at six months, against a market that keeps about half, and the difference is not what they earn. It is what they keep.

For an employer, the intervention is unglamorous and quarterly: make sure the workers you hire this quarter can settle this quarter. Managed accommodation near the plant is the most direct form of that support, and in corridors like Hosur it already exists within walking distance of the gate. The broader argument, that continuity is the infrastructure layer India's industrial growth is missing, is here: Why continuity matters. If your plant is losing its operators before day ninety, the fix is not on the payslip. It is in the room the payslip has to survive.