Cutting COD returns and RTO.

RTO is a storefront problem

Return to origin gets treated as a shipping line item, which is why it rarely gets fixed. The order comes back, the courier gets blamed, the cost gets absorbed. But the decisions that produced that failed delivery were nearly all made on your website: which payment methods you offered, how you asked for the address, whether you confirmed intent, and whether you gave the customer a reason to pay upfront.

The economics make this worth real engineering time. A returned COD order costs you forward shipping, return shipping, the working capital locked up in between, the ad spend that acquired the order, handling, and often a product that comes back unsellable. Brands typically put the all-in cost per RTO order in the mid hundreds of rupees. Against a normal D2C contribution margin, a handful of RTO orders erases the profit on a much larger number of good ones.

The numbers worth knowing

• COD dominates volume. Roughly 60 to 65 percent of orders on Indian platforms are still cash on delivery, and the share is higher outside the metros.
• COD dominates losses disproportionately. Prepaid return rates sit in the low single digits. COD return rates are multiples of that, which is why COD generates the large majority of total RTO despite being a smaller majority of orders.
• Growth is coming from where RTO is worst. Around 66 percent of new D2C orders in FY26 came from tier 2 and tier 3 cities, and those markets run materially higher RTO than metros. Your RTO rate will drift upward as you succeed, unless you actively work against it.
• It spikes at peak. The festive peak is the worst period of the year for RTO, which is worth planning into your peak season preparation rather than discovering in December.

Fix the checkout first

This is where the highest-leverage changes are, and where most brands make the mistake of reaching for the blunt instrument.

• Price the difference, do not just ask. A prepaid discount or a COD handling fee changes behaviour far more reliably than a nudge. Reported conversion is strong when the incentive is concrete and visible at the payment step rather than buried in the cart. Set the incentive below your per-order RTO cost and it pays for itself even at a modest take rate.
• Show the incentive at the moment of choice. A prepaid discount the customer discovers after selecting COD converts nobody. It belongs next to the payment options, stated as the actual rupee amount saved on this order.
• Verify the phone number. An OTP at checkout on COD orders removes a meaningful slice of fake and mistyped orders, which are pure loss. Keep it to COD, where the risk is, rather than taxing prepaid customers with an extra step.
• Consider partial advance. A small token amount on COD orders converts an intent-free order into a committed one, and brands running it consistently report lower RTO on the orders that go through.
• Do not let payment options crowd the page. Five wallets, three BNPL logos and a UPI grid is not more choice, it is more hesitation. Lead with the methods your customers actually use.

Fix the address, because undeliverable is not the customer's fault

A large share of failed deliveries are not refusals at all. They are addresses the courier could not resolve, or a phone nobody answered.

• Autocomplete the address. Typed addresses carry typos, missing localities and wrong pincodes. Autocomplete removes an entire category of failure and shortens the form at the same time.
• Validate pincode serviceability before payment, not after. Telling someone at checkout that you cannot deliver is a lost sale. Telling them after they have paid is a refund plus a bad review.
• Ask for a landmark, on mobile. It looks old-fashioned and it demonstrably improves first-attempt delivery outside metro addressing systems.
• Cut every field you do not need. Each additional field costs conversion and adds a place to make a mistake. Company name, address line 2 and alternate email are usually not earning their keep.

The post-order window is where conversion actually happens

The order is placed but not yet dispatched. This is the most underused hour in Indian D2C, and it is where COD orders get converted at scale.

• Send a payment link on WhatsApp before dispatch. A short message confirming the order, offering the prepaid discount and carrying a one-tap payment link converts a meaningful share of COD orders, with brands commonly citing 10 to 20 percent. It works because intent is highest immediately after the order and the customer is already expecting a message.
• Confirm the order, and mean it. A confirmation that asks the customer to verify the address and expected delivery window catches wrong addresses before you have paid to ship to them.
• Give an easy cancellation window. Counterintuitive, and correct. A customer who cancels before dispatch costs you nothing. The same customer refusing at the door costs you the full round trip.
• Keep them informed after dispatch. Most refusals at the door are people who forgot they ordered or were not home. Proactive delivery notifications on the channel they actually read is the cheapest RTO reduction available.

Selective COD, handled carefully

Risk scoring lets you restrict COD where it loses money without removing it where it earns. Score on the signals that actually predict RTO for your catalogue: pincode history, first-time versus repeat customer, order value against your average, product category, and whether the address resolved cleanly.

Then act proportionately. Hiding COD entirely for a high-risk segment is the strongest lever and the riskiest one. Softer options usually net out better: require a partial advance above a value threshold, add a COD fee that scales with risk, or require OTP verification for first-time buyers only.

Be careful here. In tier 2 and tier 3 markets, COD is not laziness, it is trust: it is how a customer buys from a brand they have not bought from before. Removing it wholesale in the markets driving two thirds of your new orders will cut RTO and your revenue together. Restrict at the margins, and give people a reason to prepay rather than no alternative.

Measure it properly

An overall RTO percentage is a number you cannot act on. Break it down until it points at a decision.

• By pincode and by region. The tail is usually concentrated. A small set of pincodes often accounts for a disproportionate share of failed deliveries.
• By SKU and category. Certain products attract impulse COD orders that get refused. Sometimes the fix is the product page, not the payment method.
• By acquisition source and campaign. Aggressive discount-led campaigns reliably produce lower-intent orders and higher RTO. If you are not attributing RTO back to campaigns, you are optimising ad spend against revenue that never arrived. Our GA4 and Shopify setup guide covers getting that attribution clean.
• By first-time versus repeat. Repeat customers RTO far less. This is the number that justifies retention spend to a finance team.
• As cost, not percentage. Convert RTO into rupees per order shipped and put it next to your contribution margin. It changes the conversation from a logistics metric to a P and L line, and it sets the ceiling on what a prepaid incentive is worth paying.

Common mistakes

• Going prepaid-only to solve it. It solves RTO and creates a revenue problem, especially outside metros. Very few Indian D2C brands can afford to give up COD volume outright.
• Discounting harder than the RTO costs. If your prepaid incentive exceeds your all-in RTO cost per order, you are paying to avoid a cheaper problem. Do the arithmetic first.
• Blaming the courier. Courier performance matters, and switching partners is worth doing on evidence. It will not fix orders that were never real, addresses that never resolved or customers who forgot they ordered.
• Treating it as a seasonal fire. RTO spikes at festive peak because volume and low-intent traffic spike together. The brands with the lowest festive RTO fixed their checkout and post-order flow months before the peak.

FAQ