The real cost of one returned COD order

Most merchants price an RTO at the cost of the return leg. It is the number the courier invoices, so it is the number that sticks. It is also, on almost every catalogue we have looked at, less than half of what the order actually cost.

Work through a single ₹1,800 fashion order that ships COD and comes back.

The line items

Forward freight. You already paid it. The parcel went out, travelled, and failed to convert. Call it ₹70.

Return freight. The courier charges again to bring it home, usually at the same or a slightly higher rate. Another ₹70 to ₹90.

Packaging. The polybag, the filler, the invoice, the tape. Small individually, real at volume. ₹25.

Handling on both ends. Someone picked, packed and manifested the order. Someone else received it, opened it, checked it and put it back into sellable stock — or didn't. Fifteen minutes of warehouse time is conservative, and at ₹250 an hour that is ₹60.

Payment gateway and platform fees. Usually reversed, sometimes not, occasionally with a fixed component that is not.

Working capital. The stock was unavailable for the eight to fourteen days it spent in transit. On a fast-moving SKU during a sale window, that is the expensive one — and it never appears on any invoice.

Quality loss. A meaningful share of returned apparel cannot go back to A-grade stock. Even a 10% write-down rate across returns is ₹180 amortised on this order.

The freight-only number is roughly ₹150. The honest number, including the write-down and the labour, lands between ₹380 and ₹450 — a quarter of the order value, on an order that generated no revenue at all.

What that changes

Once the number is right, three decisions get easier.

A prepaid discount stops looking expensive. A 7% incentive on this order costs ₹126. It is cheaper than the return, and it is spent only on orders that convert. The comparison is not "discount versus full margin" — it is "discount versus a probability-weighted RTO."

Verification stops looking like friction. The objection to OTP is always the same: it will cost me conversions. It will cost you some. The question is whether the orders you lose were ever going to be delivered. A shopper who will not confirm a phone number they typed thirty seconds ago is not a shopper you were about to profit from.

Rules become arithmetic instead of instinct. If your RTO rate on ₹5,000-plus COD carts is 30% and each failure costs a quarter of order value, every such order carries roughly ₹375 of embedded expected loss. That is what a COD cap above ₹5,000 is actually buying you.

Do this with your own numbers

Pull your last ninety days and calculate three figures: RTO rate, average order value, and the fully-loaded cost of a return using the line items above. Multiply them. That product is your monthly RTO bill, and it is almost always larger than the number you have been carrying in your head.

Then set your first rule against the segment where that bill concentrates — usually high-value carts, a handful of pincodes, or a single category. You do not need a policy for the whole catalogue on day one.