COD & RTO guide

How to reduce COD returns in Pakistan

Return-to-origin is the quiet tax on Pakistani e-commerce. Here's why cash-on-delivery orders bounce back, what each return really costs you, and the tactics, starting with automated confirmation calls, that bring your return rate down.

Updated July 2026

If you sell online in Pakistan, you already know the problem. Cash on delivery is how the country shops (roughly 75 to 95% of e-commerce orders are COD), and it's also where the money quietly leaks out. To reduce COD returns in Pakistan you first have to see the leak clearly: parcels that ship, travel, and come straight back unopened. That's return-to-origin, or RTO, and across the market it commonly runs about 20 to 35% of COD orders. Every one of those is your product going on a round trip at your expense.

The good news is that RTO is fixable, and the biggest single fix is boring and cheap: talk to the buyer before you dispatch. An automated order-confirmation call ("press 1 to confirm, press 2 to cancel") routinely brings a return rate from the 30s down toward roughly 18 to 22%. This guide walks through why returns happen, how much money is at stake, and how to stack a handful of tactics so fewer parcels come back.

Key takeaway: RTO comes down when you confirm the buyer actually wants the parcel before it moves, so start with a confirmation call on every order. Then add address checks, WhatsApp and advance payments on top.

Why COD returns happen

Returns feel random, but almost every failed COD delivery falls into one of four buckets. Once you know the bucket, you know which tactic fixes it.

  • Fake, prank or duplicate orders. With no payment step, anyone can place an order with a made-up name or a competitor's address. Nobody's home, or nobody wants it. These are pure loss. The parcel just tours the country and comes back.
  • Buyer's remorse. The customer ordered at 1am on impulse, spotted the same item cheaper elsewhere, or simply cooled off by the time the rider turned up two days later. The longer the gap between order and delivery, the more of this you get.
  • Wrong or incomplete address. A missing block number, an old address, a phone the rider can't reach. The courier tries once or twice, gives up, and sends it back.
  • No confirmation at all. Nobody ever checked the order with the buyer. The strongest single predictor of a return is a parcel that shipped without anyone confirming the customer actually wants it.

Notice that three of these four are decided before the rider ever knocks. That's why the confirmation step matters so much more than anything you do on delivery day.

The RTO math: what returns really cost

It's easy to shrug at a 30% return rate until you put rupees on it. Let's walk through a simple example, and plug in your own numbers as you read.

Say you ship 1,000 COD orders a month at an average order value of PKR 2,500. At a 30% RTO rate, that's 300 parcels a month coming back. Suppose each failed delivery costs you PKR 300 in two-way courier fees, plus packaging and the staff time to put it back on the shelf. That's roughly PKR 90,000 a month spent on parcels that earned you nothing, and that's before you count the cash tied up in stock that was on a truck instead of on a shelf.

Now bring RTO down from 30% to 20% with a confirmation workflow. You're at 200 returns, which is 100 fewer failed deliveries a month, around PKR 30,000 saved, plus 100 units of stock free to sell again. The calls that got you there cost a fraction of that. Market rates run about PKR 3,000 per 500 calls (roughly $0.018 to 0.02 per answered call), and if you run them through your own cloud PBX it all folds into one monthly bill.

Roman Urdu: Har order se pehle ek chhoti confirmation call (press 1 confirm, press 2 cancel) aur aap ka RTO khud ba khud neeche aa jata hai.

Tactic 1: automated order-confirmation calls

This is the one to set up first, because it works on three of the four return buckets at once. The moment an order lands, an automated call goes out to the buyer: a short recorded greeting in Urdu or English, then "press 1 to confirm your order, press 2 to cancel." Press 1 and the parcel ships. Press 2, or don't answer after a couple of tries, and it never leaves the warehouse.

That one call quietly weeds out fake orders (there's no real buyer to press 1), catches remorse (people cancel on the call instead of at the door), and flags bad numbers (a dead line is a warning sign before you've paid for shipping). Because it's automated, it scales. Whether it's 50 orders or 5,000, the calls go out the same way, and you get a clean list of confirmed orders to hand your courier each morning. See exactly how the flow works on our Shopify COD confirmation page.

Two things make voipcom's version different from a bolt-on app. First, the calls come from your own Pakistan DID number: a real Islamabad 051 line today, with Karachi 021, Lahore 042 and mobile numbers launching next. Customers recognise a proper business line and pick up, where an unknown mobile gets ignored. Second, the same IVR system that runs your confirmation calls also answers your sales and support lines, so you're not paying for a separate tool. It all lives in one cloud PBX, from PKR 3,200/month, billed in PKR, with no hardware. Compare the tiers on our pricing page.

More tactics that stack on top

Confirmation calls do most of the work, but the stores with the lowest RTO add a few more habits on top. None of these replace the call. They close the gaps it leaves.

  • Address & number verification. On the confirmation call, or in a quick follow-up, read the address back and confirm a working phone number. A five-second check puts a stop to the "rider couldn't find it" returns.
  • WhatsApp confirmation. For buyers who never pick up, a WhatsApp message with a confirm/cancel reply is a cheap second net. Pair it with the call rather than relying on it alone, because a lot of chats go unread.
  • Advance or partial payment on high-value orders. Ask for a small deposit or full prepayment on expensive items. It filters out non-serious buyers and shares the shipping risk. Keep full COD on low-ticket products, where the extra step would cost you more sales than it saves.
  • Blacklist repeat returners. Keep a simple list of numbers that have refused delivery before. When one of them orders again, ask for prepayment or skip the order. A handful of chronic returners can quietly wreck your average.
  • Ship faster. Remorse grows with every day the parcel is in transit. Confirmed orders that go out the same day come back far less often than ones that sit for a week.
Which tactic fixes which return
TacticBest forReturns it stops
Confirmation call (press 1 / 2)Every orderFake orders, remorse, dead numbers
WhatsApp confirmationBuyers who don't pick upRemorse, unreachable buyers
Address & number checkEvery orderWrong / incomplete address
Advance / partial paymentHigh-value ordersNon-serious & fake buyers
Blacklist repeat returnersKnown problem numbersSerial refusals

Putting it together

You don't need all five tactics on day one. The order that works: turn on confirmation calls for every order, verify the address on that same call, and start a blacklist. Then add WhatsApp as a backup and advance payments for your most expensive lines. Check your RTO every week. That number becomes your scoreboard, and most stores see it move within the first month.

If you're still juggling personal SIMs and a spreadsheet of orders, the confirmation call is also a good reason to move to a proper cloud PBX: one business number, every call logged and recorded, and the same system confirming orders and answering customers. New to the jargon? Start with what a PBX is, or book a free demo and we'll run a live confirmation call from start to finish.

FAQ

Common questions about reducing COD returns

How do I reduce COD returns in Pakistan?

Start with an automated confirmation call on every order (press 1 to confirm, press 2 to cancel) before you hand the parcel to the courier. Then add address verification, a WhatsApp confirmation, an advance or partial payment on high-value orders, and a quiet blacklist of repeat returners. The confirmation call does most of the work, because it catches fake orders and buyer's remorse before the courier has moved.

What is a good RTO rate for Pakistani e-commerce?

Return-to-origin on cash-on-delivery orders in Pakistan commonly runs about 20 to 35%. A confirmed-order workflow can bring that down toward roughly 18 to 22%. If yours sits above 30% for long, it's eating your margin and is worth fixing first.

Do order confirmation calls actually reduce returns?

Yes. A short automated call asking the buyer to press 1 to confirm or 2 to cancel removes fake and duplicate orders, catches buyer's remorse, and fixes wrong addresses before dispatch. In the Pakistani market it's the single most reliable way to bring RTO down toward roughly 18 to 22%.

How much do COD confirmation calls cost in Pakistan?

Market rates for standalone confirmation-call apps are roughly PKR 3,000 per 500 calls, or about $0.018 to 0.02 per answered call. With voipcom the confirmation calls run through your own cloud PBX and Pakistan DID number, so the same automation sits alongside your regular business calls from PKR 3,200/month, billed in PKR, with no hardware. See pricing.

Should I ask for an advance payment on COD orders?

For high-value orders, yes. A small advance or partial payment through a wallet or bank transfer filters out non-serious buyers and shares the shipping risk. Keep full COD for low-ticket items, where the extra step would cost you more sales than it saves.

How do I stop fake COD orders?

Confirm every order by voice or WhatsApp before dispatch, verify the address and a working phone number, and keep a blacklist of numbers that have refused delivery before. Fake and prank orders almost never survive a confirmation call, because there's nobody on the other end willing to press 1.

Confirm every order before it ships

Book a free demo and we'll run a live "press 1 to confirm" call on your own catalogue, then show you the confirmed-order list your courier gets each morning.