Attrition In Indian Ecommerce Ops: The Real Cost

Most Indian ecommerce operations do not have an attrition problem they can see. They have a replacement cost that has never been calculated, absorbed quietly into overtime, error rates and supervisor time.

What attrition actually looks like at this level

Blue-collar attrition in India runs at 50 to 80 percent a year in warehouse and logistics roles. Compare that with the all-India figure across sectors of roughly 17 percent in 2025, projected closer to 14 percent for 2026, and it becomes clear why a white-collar HR playbook applied to a fulfilment floor achieves nothing.

Two sharper numbers matter more than the annual figure:

• Roughly 30 to 40 percent of new warehouse hires leave within the first three months. Your problem is almost always the first ninety days, not the second year.
• Average tenure in these roles sits near twenty-one months, and about 69 percent of blue-collar hires are temporary or contract, with low conversion to permanent.

So stop reporting annual attrition as a single number. Report 90-day survival rate, and report it by supervisor. The variance between two supervisors on the same floor is usually wider than the variance between two cities.

The true replacement cost

Take a picker on Rs 16,000 a month. The visible cost of replacing them is a staffing vendor fee or a referral bonus. The real cost has six parts.

• Sourcing: Rs 2,000 to Rs 5,000 through a staffing vendor, or Rs 1,000 to Rs 2,000 as a referral payout.
• Onboarding admin: documentation, PF and ESI enrolment, ID, medical, uniform and PPE. Roughly Rs 1,500 to Rs 2,500.
• Trainer time: five days of a senior person’s attention, even if only partial. Call it two full days of a Rs 25,000-a-month supervisor.
• Trainee wages during non-productive training: five days at the joiner’s own rate.
• Ramp loss: a new picker runs at 50 to 60 percent of standard rate for the first two to three weeks. That gap is output you paid for and did not receive.
• Error cost: the elevated mis-pick and mis-pack rate over the first sixty days, at roughly Rs 250 to Rs 600 per incident.

Added together, replacing one Rs 16,000 picker costs Rs 25,000 to Rs 35,000, which is about 1.5 to 2 times monthly wage. A forty-person warehouse running 60 percent attrition replaces twenty-four people a year. That is Rs 6 lakh to Rs 8.5 lakh, most of it never appearing as a line item anyone owns.

Why people actually leave

Exit interviews at this wage level are close to useless, because the answer is always personal reasons or better salary. The real drivers show up when you look at who leaves and when.

• The salary credit date moves. Someone budgeting weekly cannot absorb a salary that lands on the 7th one month and the 11th the next. This is the most under-rated driver in Indian operations.
• The roster is published late. A person who learns their shift the evening before cannot arrange childcare, a second income or a shared commute.
• Commute cost and the last shift. If the shift ends at 10pm and there is no transport, the job carries a hidden cost of Rs 60 to Rs 100 a day.
• The supervisor. Most resignations at this level are resignations from a person, not from a company.
• Basic conditions. Cool drinking water, working toilets, a rest area that is not the loading dock. The absence of these reads as a signal about how the employer sees them.
• No visible next step. Picker today, picker in two years, same pay band, no title change.

Retention levers that work at this wage level

Several of these cost nothing. The ones that cost money cost less than replacement.

• Pay on the same date every month without exception, and send an SMS confirming the credit. Reliability beats a Rs 500 increment.
• Publish the roster seven days ahead and hold to it. A change should require the supervisor to ask, not to inform.
• Pay continuity bonuses rather than joining bonuses. Something like Rs 1,000 at day 90, Rs 2,000 at month six, Rs 5,000 at month twelve. Joining bonuses buy attendance for a fortnight.
• Run referrals properly. Referral hires from a worker’s own community retain measurably better. Pay Rs 1,000 to Rs 2,000, half at joining and half at day ninety, so the referrer stays invested in the outcome.
• Build a three-step ladder with real titles and real pay gaps: picker, senior picker or QC, team lead, with Rs 1,500 to Rs 2,500 between steps and a written qualification for each.
• Provide transport for any shift ending after 8pm, or a fixed conveyance allowance for it.
• Run structured check-ins at day 7, day 30 and day 60, conducted by someone other than the direct supervisor. Most 90-day exits are visible at day 30 if anyone bothers to ask.
• Train supervisors. One day on giving clear instructions, correcting errors without humiliation, and escalating grievances pays back faster than any incentive scheme.

How to know it is working

Put three numbers on the monthly operations review and give each one an owner. Ninety-day survival rate. Attrition by supervisor. Cost of replacement, calculated with the six components above and stated in rupees.

Then set a target that is achievable rather than aspirational. Moving 90-day survival from 65 percent to 80 percent at a forty-person site removes roughly six replacements a year and around Rs 1.8 lakh of direct cost, before you count the error reduction that comes from having a more experienced floor. That is the argument to take to finance, and it lands harder than a slide about engagement.

One last point on framing. This workforce is not casual about work. People take these jobs seriously and leave for concrete, rational reasons. Treat attrition as an operations defect with measurable causes, the same way you would treat a spike in mis-picks, and the fixes become obvious.