Is My Ecommerce RTO Rate Normal — and What Is It Costing Me?

For an Indian D2C brand, normal is 29% RTO on COD orders and 6% on prepaid. About 19% blended at a 55% COD share at ₹1,200–1,800 AOV. Each return costs ₹210 to operate and about ₹670 once wasted ad spend is counted. The rest of this article turns that answer into decisions.

The numbers here come from 50+ Indian D2C brands and roughly 156,000 shipped orders analysed by Profitbox360 between January and July 2026 — COD-heavy brands at ₹1,200–1,800 AOV.

What RTO rate is normal for an Indian D2C brand?#

Payment mode decides most of it: 29% on COD, 6% on prepaid. Category decides the rest. Fashion runs around 45%, beauty around 17%, electronics around 20%. Judge your number against your mix, not one industry average.

The blended math at 55% COD share: (55% × 29%) + (45% × 6%) = about 19% overall. A brand quoting 20% RTO at 80% COD share is doing better than one quoting 15% at 30% COD. The COD share tells you more than the headline number.

Why the spread: fashion is bought on impulse, sized wrong, and refused easily. Beauty is a small, wanted parcel — easy to accept. Electronics sit in between: considered purchases, but a big amount at the door still gets refused. Brands that run the full reduction ladder — verification, address quality, partial COD — land their COD RTO around 15%. That is what good looks like against a 29% baseline.

What does one RTO actually cost — and who pays it?#

₹210 in direct operating cost. The seller pays all of it. The buyer who refused at the door pays nothing. Count the ad money that bought the order and the true cost is about ₹670 — and ₹780 if the order came from ads.

The stack, line by line:

One line is missing on purpose. The ₹40 COD handling fee is not in the stack — couriers reverse it when a shipment comes back. It is a cost of delivered COD orders, not of failed ones. Any stack that includes it double-counts.

Two lines deserve a second look. The ₹460 of ad money was already paid for a buyer who returned nothing — and it is more than the ₹400 an average order carries, because returns skew toward ad traffic: orders from ads fail at twice the organic rate, so about four of five returns carry the full ~₹570 paid-order ad cost. And below the table sits a hidden line: about 5% of returned units come back unsellable. At 1,865 monthly returns, that is roughly 93 units written off. At a ₹420 landed cost, another ₹39,000 a month nobody budgets.

What is your RTO rate bleeding every month?#

At 10,000 orders a month: about ₹3.9 lakh in operating cost, ₹12.5 lakh counting wasted ads. Each point of COD RTO is worth ₹11,500 to ₹31,000 a month depending on how you remove it. Price every fix against that.

The chain: 5,500 COD orders × 29% = 1,595 returns. 4,500 prepaid × 6% = 270 returns. Total 1,865 × ₹210 = ₹3,91,650 a month. At the ₹670 true cost: ₹12,49,550.

One point of COD RTO is 55 orders. Its value depends on what the fix does. A fix that turns returns into deliveries — better addresses, confirmation, an advance — earns the ₹360 margin and saves the ₹210 cost: 55 × ₹570 = about ₹31,000 a month. A fix that just blocks risky orders saves the cost but the sale never happens: 55 × ₹210 = about ₹11,500 a month. Rescuing an order is worth nearly three times blocking one. Prefer the fixes that deliver.

Before pricing fixes, know what is actually fixable. RTO splits four ways by cause:

One measurement warning: keep RTO separate from customer returns in your reporting. Most dashboards mix the two into one returns bucket. That hides the refusal problem and makes the fix unmeasurable. If your platform cannot split them, tag RTOs by hand. You cannot manage a 35% refusal share you cannot see.

Where does RTO concentrate — and does blocking pincodes pay?#

The worst ~5% of pincodes produce about half of all RTOs, with tier-2/3 and rural addresses over-represented. That concentration makes pincode rules the highest-leverage geography play. But blocking COD still pays much later than most brands think: the threshold is about 50% pincode RTO, not 30%.

Here is the equation everyone skips — both branches of it. Serving a COD buyer in a bad pincode earns (1 − r) × ₹360 on deliveries and burns r × ₹210 on returns. Going prepaid-only does not earn zero: about 20% of COD buyers switch and pay upfront, worth 20% × ₹363 ≈ ₹73 per buyer. COD wins while 360 − 570r > 73. Solve it: r ≈ 50%*. At 40% RTO, 100 COD buyers earn ₹21,600 on the 60 delivered and burn ₹8,400 on the 40 returned — ₹13,200 ahead. Prepaid-only keeps just the switchers: about 20 × ₹363 ≈ ₹7,300. Serving still wins by ₹5,900.

The pincode playbook, in order. Below 30% RTO: serve normally. Between 30% and 50%: do not block — raise commitment instead. Advance required, verification required, prepaid incentivised. Above ~50% observed RTO (a real population in the worst pincodes, given how hard the bad 5% skews): prepaid-only finally wins — and even there a fifth of the buyers stay and prepay. Blocking below the threshold throws away profitable orders.

What should your COD cap be — and which orders shouldn't get COD at all?#

Cap by order value. Full COD below ₹1,500. COD with verification or a partial advance from ₹1,500 to ₹3,000. An advance or prepaid-preferred above ₹3,000. The chance of refusal is roughly the same on every order. The cost of a refusal grows with order value. The caps just match the rule to the cost.

(Marketplaces cap COD around ₹50,000. That cap is for fraud, not for you. Your math kicks in far earlier.)

Beyond value, three order types should lose COD at any ticket. Buyers with a past RTO against their number — the strongest refusal predictor in your data. Custom or made-to-order items, where a return is a write-off, not restock. And pincodes above the 50% threshold from the previous section. COD is a conversion tool. Offer it where the math works, add conditions where it is marginal, withdraw it where it fails.

FAQ#

Is 20% RTO normal in India? On COD orders, 20% is better than the 29% baseline. Blended across payment modes at a typical COD share, it is close to the ~19% norm. Always say which one you are quoting.

Who pays RTO charges — the seller or the courier? The seller, both legs. Forward and reverse shipping are billed to you; the COD fee is reversed when the shipment comes back. The buyer who refused pays nothing. That is why stopping an RTO before shipping beats everything you can do after it ships.

What is the average RTO rate in Indian ecommerce? About 19% blended at a 55% COD share. 29% on COD orders, 6% on prepaid. Category swings the COD number from ~17% (beauty) to ~45% (fashion).

How is RTO different from a customer return? An RTO never reached the buyer — refused, unreachable, or undeliverable. A return was delivered and then sent back. Different causes, different fixes. Report them separately.

Do high-RTO pincode lists work? As a routing signal, yes — the worst 5% of pincodes drive about half of all RTOs. As a block list, mostly no. Blocking pays only above about 50% pincode RTO — and blocking means prepaid-only, which still keeps the fifth of buyers who switch. Below that, raise commitment — advance or verification — instead of refusing service.