RTO or customer return? The answer changes your fix.
One is a delivery problem, the other is a listing problem. They cost different amounts, need different evidence and respond to opposite remedies — yet almost every seller tracks them as a single number.
- RTO = delivery never completed. Customer return = buyer received it, then sent it back.
- RTO is a delivery-side problem: COD exposure, address quality, courier performance.
- Customer returns are a listing-side problem: sizing, description accuracy, expectation gaps.
- Different claim routes — charge accuracy for RTO, goods condition for customer returns.
- Robnu classifies each return automatically so you can see which half is costing you. Free while we figure out pricing.
A combined returns percentage is one of the least useful numbers in marketplace operations. It tells you money is leaving without telling you why, and the two underlying causes need almost opposite responses.
Did delivery complete?
That is the whole classification. If the parcel never reached the buyer’s hands — refused, undeliverable, unavailable across attempts — it is an RTO. If the buyer took delivery and later sent it back, it is a customer return.
The consequence that matters most is evidence. In an RTO nobody opened the parcel, so condition disputes essentially cannot arise — your claims are about whether the freight charge was correct. In a customer return the parcel has been opened, and what comes back may not be what went out.
Two problems, compared
Everything below differs between the two — which is the argument for tracking them apart.
RTO
Cause: COD refusal, bad address, customer unavailable.
Cost: freight both ways, lost margin, stock in limbo.
Claim: charge accuracy — weight, duplicates, never returned.
Fix: prepaid conversion, address validation.
Customer return
Cause: sizing, expectation mismatch, quality, buyer changed mind.
Cost: freight plus product risk — used, swapped or missing goods.
Claim: goods condition — needs video evidence.
Fix: listing accuracy, sizing, photos.
Two rates, tracked separately, automatically
Splitting returns by hand means checking, for every single return, whether delivery completed before the return began — then maintaining two running figures. It is simple and nobody sustains it, which is why the combined number persists.
Robnu is an agentic OMS. It classifies every return by what actually happened, tracks both rates independently and by SKU, and routes each claim down the correct path with the right evidence attached — a rare approval click while fully-autonomous filing rolls out. You stop guessing which problem you have.
You sell. Robnu runs the rest — and makes sure every rupee is paid correctly.
RTO vs customer return, answered
An RTO happens before the customer takes possession — delivery failed, was refused, or the address was unreachable, so the courier sends the parcel back. A customer return happens after successful delivery, when the buyer decides to send the item back. The dividing question is simply whether delivery ever completed, and that single fact changes almost everything downstream.
Because the causes, costs and remedies are entirely different. RTO points to delivery-side problems: cash-on-delivery exposure, address quality, courier performance. Customer returns point to product-side problems: sizing, description accuracy, expectation mismatch. Fixing one does nothing for the other, so tracking them together means you cannot tell which to work on.
It depends on category and that is exactly why you need them separated. RTO is freight-heavy — you pay both legs with no revenue. Customer returns add product risk on top: the item comes back opened, possibly used, possibly not even the right item. For higher-value goods the condition risk usually outweighs the freight.
Yes, and using the wrong one is a common reason claims fail. RTO claims are about charge accuracy — wrong weight, duplicate deduction, parcel never returned. Customer return claims are about goods condition and require evidence captured at the moment of opening. The evidence that wins one is largely irrelevant to the other.
Not in the strict sense, but the reverse leg of a customer return can fail to reach you, which produces a similar-looking outcome: a return you were charged for that never arrived. That is a lost-in-transit claim rather than an RTO, and it is worth catching because you lose stock and freight together.
For RTO: shift toward prepaid, validate addresses before dispatch, and improve packaging so parcels are not refused on sight. For customer returns: accurate sizing charts, honest photography, precise material and dimension descriptions. The first set is operational, the second is about the listing — almost no overlap.
Tag every return with whether delivery completed before the return began. That one field cleanly splits the two and lets you compute both rates independently. Most sellers keep a single combined returns percentage, which is precisely why the underlying cause stays invisible.
It is always expensive, but the response depends entirely on the split. A high rate driven by RTO in cash-on-delivery orders is a distribution problem. The same headline rate driven by customer returns on one SKU is a listing problem with a specific fix. The number alone tells you almost nothing actionable.