Read first
The order placed at 1 am
A COD order comes in at 1 am for a ₹799 product, with a phone number that looks like it was typed by a cat and an address that says "near temple, main road". The owner ships it because an order is an order. Four days later the courier phones a number that does not exist and the parcel starts its journey home. You paid for both journeys, and the buyer never knew.
Cutting RTO is mostly about not shipping that order, and about saving the real ones that nearly fail. Start with what is cheapest and do one lever at a time so you can see which one worked.
The levers, in the order to try them
| Lever | What you do | Cost | Watch out for |
|---|---|---|---|
| 1. Address and phone at checkout | Require pincode check, a full address with landmark, and a phone number validated by OTP | Low: a checkout app or plugin | Extra steps lose some real buyers; test the drop |
| 2. Confirm COD before dispatch | Send a WhatsApp or IVR message asking the buyer to confirm. No reply in a set time means hold or cancel | Message cost, or an app fee | Slows dispatch by hours; tell buyers at checkout |
| 3. Prepaid nudge | Show a small discount or free shipping for UPI or card at checkout | The discount you offer | Do not hide COD; fraud checks still apply to prepaid |
| 4. Act on the first failed attempt | Call the buyer the same day a courier reports a failed delivery; reschedule | A person's time or an automated flow | Your courier panel must let you give new instructions |
| 5. Pincode and order-value rules | Turn COD off, or ask for a token advance, in pincodes or order bands where your RTO is high | Some lost orders | Do it from your data, not from rumour |
| 6. Courier by pincode | Compare couriers per pincode and route each region to the one that delivers there | Time to set up in your aggregator | A courier cheap in one city may be poor in the next |
| 7. Product and promise | Ensure the buyer knows what they are getting: clear photos, size guide, price including delivery | Content time | A surprise at the door is a refusal |
Why the first failed attempt matters most
One logistics network published how parcels end up depending on how many delivery attempts failed: 20.2% after one; 45.5% after two; 88.5% after three or more. In other words, after one failed attempt most parcels still reach the buyer, after two about half come home, and after three or more almost all do. The value of a same-day call after the first failure is therefore the highest-return action you have in operations. See what an NDR is and what to do.
RTO is also uneven by place. One network's data for October and November 2025 gave all-payment RTO of Metro 6.6%; tier 1 8.1%; tier 2 11.5%; tier 3 and beyond 12.0%. If most of your customers are in tier 2 and 3 towns, expect a higher base and invest earlier in address checks and calls.
COD confirmation, step by step
- At checkout, tell the buyer "We will message you on WhatsApp to confirm your COD order."
- When the order arrives, send a message with the items, total and a one-tap confirm button, or a short call with an IVR prompt.
- If the buyer confirms, ship. If there is no reply within a set time, such as 6 to 12 hours, send one reminder, then hold the order.
- If the buyer cancels, cancel and release stock. This is a cheap cancellation, not an RTO.
- Record confirm rate and RTO of confirmed versus unconfirmed orders. After a month you will know what the check is worth.
Keep the message about the order itself: items, total, delivery address. Messaging a customer about offers is a different category on most WhatsApp providers' rate cards and needs the buyer's permission, so check your provider's rules before you add anything promotional.
What usually goes wrong
- Turning on every lever at once and not knowing which did anything.
- Offering a prepaid discount so large it costs more than the RTO it prevents. Compare it with the ₹150–300 per order cost of a returned order and with your COD fee.
- Banning COD overall, losing the buyers who only pay on delivery and giving the sale to a rival. Restrict by data, not by habit. Vendor estimates put COD at 50–65% of orders in D2C, so a brand cannot always switch it off.
- Confirming by call at 2 am. Use hours when people answer.
- Ignoring repeat offenders. Keep a list of numbers and addresses with a past RTO and decide a rule for them.
Fake and prank orders are a separate problem from honest non-delivery. They need blocking and verification, not rescue calls.
Apply it to your store
Pick the one lever you can switch on this week. Measure COD RTO for the 30 days before and after, using the method in what a normal RTO rate is. If your rate stays high after levers one to four, look at the product and the audience before the courier. Tell us where you are and we will go through your last month with you.
Questions people ask
Should I stop offering COD?
Not by default. COD is about half or more of D2C orders by vendor estimates, and many buyers will not pay in advance. Restrict it by pincode, order value or buyer history from your own data first.
Does a prepaid discount lower RTO?
Vendors say it moves some buyers to prepay, and prepaid RTO is far lower than COD in every published source. The size of the effect depends on your product, so test it and compare its cost with the cost of an RTO.
How soon should I call after a failed delivery attempt?
The same day if you can. Published network data shows most parcels still deliver after one failed attempt but very few do after three.
Where these numbers come from
- Share of orders that end as RTO after one, two and three or more failed delivery attempts: 20.2% after one; 45.5% after two; 88.5% after three or more (all India; ClickPost network). Reported: ClickPost (checked research 2026-11-14), checked 13 November 2026.
- RTO by city tier, all payment modes: Metro 6.6%; tier 1 8.1%; tier 2 11.5%; tier 3 and beyond 12.0% (all India; ClickPost, Oct to Nov 2025, 35.8 million orders). Reported: ClickPost (checked research 2026-11-14), checked 13 November 2026.
- What one RTO order costs: ₹150–300 per order (all India; Razorpay buffer figure; ClickPost models ₹350 on a ₹1,200 order (range ₹285 to 490)). Reported: ClickPost (checked research 2026-11-14), checked 13 November 2026.
Fees, rates and rents change. Check the official source before you pay or sign anything.
Operations · Product brands
More on operations for a product brands
16 pages on this. You are on number 4.
- What is a normal RTO rate for a brand shipping by courier in India, and what is a bad one?
- White label, private label and contract manufacturing: which one am I buying?
- How to cut RTO on cash-on-delivery orders for your brand
- How to find a good contract manufacturer for your brand in India
- What can your brand do with slow-moving or dead stock?
- How to check a factory's quality before you pay for your brand's first batch
- How to plan stock and cash for Diwali, wedding and sale seasons in your brand
- Which courier or shipping aggregator should your brand use in India?
and 8 more.
Reading about it is one thing; doing it in your product brands is another. If you'd like a hand running it smoothly day to day, tell us where you are and a person will reply, free.
Get help with this ↓What now?
Where are you?
One tap, and we point you to the right next step.
What to read next
Related questions
For your kind of business
- D2C brandHow do I block fake and prank COD orders on my D2C store?→
- D2C brandHow should an Instagram or WhatsApp brand handle COD risk?→
- D2C brandHow do I set up COD verification by OTP or WhatsApp?→
- D2C brandHow do I reach customers in tier 2 and tier 3 towns?→
- D2C brandShould my D2C brand stop offering cash on delivery?→