Cash on delivery is not going away in Pakistan. It carries the large majority of online orders — commonly cited above 80% — because it solves a trust problem that no amount of design will solve on its own. Customers pay when they can see the product.
The cost sits on the other side. Because a COD order requires nothing from the buyer at the moment of ordering, refusing it costs them nothing either. Published return-to-origin figures for Pakistani e-commerce range from around 18–20% at the optimistic end to 30–45% across much of the market. Every one of those parcels is billed to you twice: outbound and back.
Why the number gets so high
Three patterns account for most of it. Impulse orders placed at 1am and regretted by the time the rider arrives. Address and contact failure — a number that is switched off, an address that a rider genuinely cannot find, a customer who has left for work. And bad-faith orders, placed with no intention to accept, which are rare individually and expensive collectively.
Notice that only the last one is malicious. The other two are process problems, which is good news, because process problems are fixable.
Confirm the order before it ships
The single most effective change is also the least sophisticated: contact the customer before the parcel leaves. A short WhatsApp confirmation with the item, the price and the delivery window, requiring a reply, removes a meaningful share of impulse orders before they cost you freight.
Automate the message, not the judgement. A first-time customer ordering a high-value item deserves a phone call, not a template.
Make the address usable by a rider
Pakistani addresses are written for humans who already know the area. A form field labelled “Address” invites exactly the kind of entry a courier cannot act on.
Break it up: city, area or sector, then street and house, then a free-text landmark field — and label the landmark field as such, because “near the mosque, blue gate” is genuinely what gets a parcel delivered. Validating the phone number at checkout and sending a confirmation SMS to it catches typos while the customer is still on the page.
Reduce the reasons to refuse at the door
A significant share of refusals are not about intent at all — the customer expected something different. Accurate product photography, honest sizing information and clear delivery timing all reduce the gap between what was imagined and what arrives.
This is the part most sellers underinvest in, because it looks like a marketing cost rather than a logistics one. It is both.
Track it by segment, not in aggregate
A single site-wide RTO percentage tells you almost nothing. Broken out by city, by product, by order value and by traffic source, it usually tells you a great deal — that one campaign produces orders that never get accepted, or that a particular product is being returned because the listing oversells it.
You cannot fix an average. You can fix a segment.
Give people a reason to prepay
You will not eliminate COD, and you should not try. But a modest discount for prepayment via card, JazzCash or Easypaisa shifts a slice of orders into a category with near-zero RTO, and improves your cash position at the same time. Even a small conversion here compounds.
Where the technology fits
None of this works if it lives in a spreadsheet. Confirmation flows, address validation, courier booking and segment-level reporting need to be built into the store itself — which is what we mean when we say an international template rarely survives contact with this market.
We build these in as standard on e-commerce projects, and we have retrofitted them onto plenty of stores that launched without them. If returns are eating your margin, tell us what your numbers look like.
