A returned COD parcel is the most expensive thing that can happen to a small Indian store, and it is expensive in a way that hides from you. There is no refund line. The money simply never arrives, and the courier invoice quietly goes up.
Let us start by putting a number on it, because most sellers have never done that.
What one RTO actually costs
A ₹900 order, 500 g, zone C, cash on delivery, refused at the door.
| Line | Amount |
|---|---|
| Forward shipping | ₹78 |
| COD handling fee | ₹30 |
| Return shipping | ₹78 |
| Packaging, unrecoverable | ₹28 |
| Your time handling the return | ₹40 |
| Product condition loss (average across categories) | ₹45 |
| Total cost of one RTO | ₹299 |
Now the part that hurts. If your contribution margin is ₹400 per order, one RTO wipes out the profit on roughly three-quarters of a successful order. At a 12% RTO rate — not unusual for an unconfirmed COD store — you are losing about ₹36 on every COD order you accept, before anything else.
Why parcels come back
In rough order of frequency, from what we see:
Note how many of those are solved by a conversation before dispatch rather than by a policy.
The eleven things that actually move the number
Confirm every COD order before you ship it
One WhatsApp message: item, price, address, and "reply YES to confirm". Unconfirmed after two attempts, cancel it. This single habit typically halves RTO. It is also the single most tedious thing in a small store, which is why most people stop doing it by order sixty.
Validate the address at checkout
Pincode against serviceability, phone number format, and a required flat/house number field. Reject an address with fewer than 15 characters — it will fail.
Offer a small prepaid discount
₹30–₹50 off, or free shipping on prepaid. Converting 20% of COD orders to prepaid is worth far more than the discount costs, because prepaid RTO is a fraction of COD RTO.
Set a COD ceiling
Above roughly ₹2,000, refusals rise sharply — the buyer needs cash on hand. Prepaid-only above your ceiling, with a clear reason shown.
Message before the delivery attempt
"Your parcel is out for delivery today, the courier will call from this number." Cuts the nobody-home bucket noticeably.
Ship faster
RTO correlates with time in transit. Every extra day is another day for the impulse to fade. Same-day or next-day dispatch is a returns strategy, not just a service promise.
Block repeat refusers
Keep a list of phone numbers that have refused twice. Offer them prepaid only. Around 2–4% of buyers generate a disproportionate share of refusals.
Match photos to reality
Over-styled photography raises orders and raises refusals. Include one plain, unflattering, accurate shot. It costs you some clicks and saves you parcels.
Choose the courier by pincode performance, not price
Delivery success varies wildly by partner and region. If your aggregator shows RTO by courier, use it. A partner that is ₹8 dearer and 4% better on delivery is cheaper.
Make cancellation easy before dispatch
A buyer who can cancel in one tap does not refuse at the door. Cancellation costs you nothing; refusal costs you ₹299.
Reconcile COD remittance weekly
Money collected at the door reaches you days later, and shortfalls happen. Nobody catches this at the end of the quarter.
Should you just turn COD off?
Depends entirely on your category and price point.
| Situation | Verdict |
|---|---|
| Average order value under ₹700, tier 2/3 buyers | Keep COD. Losing those orders costs more than the RTO |
| Average order value ₹700–₹2,000 | Keep it, with confirmation and a prepaid incentive |
| Above ₹2,000 | Prepaid only, or COD with partial advance |
| Fragile or made-to-order goods | Prepaid only. An RTO on a custom item is a total loss |
| Categories with high impulse (fashion, novelty) | Keep, but confirm ruthlessly |
Sellers who switch COD off entirely usually see orders drop 25–40% and then quietly switch it back on. The better move is almost always to keep COD and get serious about confirmation.
Tracking the number properly
You cannot improve what you do not measure, and most stores measure this wrong. Use:
RTO rate = parcels returned undelivered ÷ parcels dispatched
Calculated monthly, split by COD vs prepaid, and by courier. If you only have one number, you cannot tell whether your problem is buyers, addresses or a bad partner in one region.
A healthy target: under 5% blended, under 8% on COD alone. If you are above 15% on COD, the problem is almost certainly that you are not confirming orders.
Frequently asked questions
What is a normal RTO rate in India?
Blended across COD and prepaid, 5–10% is typical. COD alone runs 8–20% depending on category and whether orders are confirmed. Prepaid RTO is usually under 3%.
Does asking for confirmation lose me orders?
A few, and they are mostly the orders that would have been refused anyway. Sellers who start confirming typically see dispatched volume fall slightly and delivered volume rise.
Can I charge a fee for cash on delivery?
You can, and some sellers add ₹30–₹50. It suppresses COD selection, which is partly the point. Be upfront about it at the cart, not at checkout — a surprise fee at the last step costs you the whole order.
What do I do with returned stock?
Inspect, repack in fresh packaging, and put it back in sellable inventory if it is genuinely unmarked. Keep a separate bin for opened returns and discount them honestly rather than shipping them as new.
Is partial COD worth setting up?
For high-value orders, yes. Collecting ₹200 up front changes the psychology completely and filters out casual orders. It needs gateway support and a clear explanation at checkout.