Ask a struggling Pakistani online store what the problem is and you will usually hear "we need more traffic". Look at the numbers and the problem is almost never traffic. It is that the store is selling at a margin that cannot survive its own return rate, and the owner has not done the arithmetic.
Cash on delivery is the reason. It is also, for now, unavoidable for most consumer categories here. So the job is not to wish it away. The job is to understand exactly what it costs you and to systematically reduce it.
The short version
- COD is not a payment method, it is a free option you grant every customer. They can decline at the door with no cost to themselves. You pay for the delivery, the return leg and the handling either way.
- Returns come out of gross margin, not revenue, which is why a return rate that sounds survivable often is not. Run the worked example below with your own numbers.
- The courier's cash cycle is a working capital problem. You buy stock now and receive the cash days or weeks after delivery. Growth makes this worse, not better.
- Five levers move the return rate materially: confirmation, partial advance, address quality, delivery communication, and a repeat-refuser blocklist.
- Shifting even a minority of orders to prepaid changes the economics more than any advertising improvement. Make prepaid the better deal, visibly.
The arithmetic nobody runs
Take an illustrative order. These are not market figures, they are a worked example: put your own numbers in the same shape.
Suppose you sell an item for 3,000 rupees. It costs you 2,000. Your gross margin is 1,000.
Now add reality. Forward delivery costs you 250. On a refused order you also pay the return leg, say another 200, plus handling and the cost of the item being out of stock while it travels for a week.
| Scenario | 100 orders | Delivered | Refused | Gross margin earned | Delivery and return cost | Net |
|---|---|---|---|---|---|---|
| 10% refusal | 100 | 90 | 10 | 90,000 | 100 x 250 plus 10 x 200 = 27,000 | 63,000 |
| 25% refusal | 100 | 75 | 25 | 75,000 | 100 x 250 plus 25 x 200 = 30,000 | 45,000 |
| 40% refusal | 100 | 60 | 40 | 60,000 | 100 x 250 plus 40 x 200 = 33,000 | 27,000 |
Look at what happens between the first row and the third. Revenue fell by a third. Net contribution fell by well over half. And none of this includes your advertising cost, which you paid on all 100 orders including the 40 that came back.
Add a realistic customer acquisition cost and the third row is not a business with thin margins. It is a business losing money on every order, funded by stock it has already paid for.
What this should change: the return rate is not a logistics annoyance, it is the single most important number in a COD business. If you do not know yours this week, that is the first thing to fix. Our five numbers every Monday argument applies exactly here.
Five levers that actually reduce refusals
1. Confirm the order before it ships. A WhatsApp confirmation with the item, the price and the expected delivery window, requiring a reply. Orders that are never confirmed are dramatically more likely to be refused. This is the highest-return single change available to most stores and it costs one person's time.
2. Ask for a partial advance on high-risk orders. Not on everything, on the ones that fit your risk profile: high value, new customer, an area with a poor delivery record, or an order placed at 3am. Even a small advance converts a casual impulse into a commitment.
3. Fix address quality at the point of entry. A large share of failed deliveries are failures to find the customer, not refusals. Structured address fields, a required landmark, a verified phone number and a city dropdown rather than free text. Cheap to build, immediately effective.
4. Communicate the delivery window. Most refusals at the door are "I am not home" or "I forgot I ordered this". A message the morning of delivery solves a meaningful share of both.
5. Keep a blocklist and use it. A small number of numbers and addresses generate a disproportionate share of refusals. Track them. Require prepayment from repeat refusers. Almost nobody does this and it is free.
Make prepaid the better deal
You cannot remove COD in most categories. You can make prepayment visibly more attractive:
- A discount for paying online that is smaller than what a refusal costs you. That is the entire calculation.
- Free delivery on prepaid orders only.
- Faster dispatch for prepaid, and say so on the product page.
- Payment that takes ten seconds, not a bank transfer with a screenshot. Instant rails make this genuinely easy now: see Raast, QR codes and the end of the cash-only business.
A store that moves from almost entirely COD to a meaningful prepaid share has changed its unit economics more than any creative refresh or bid adjustment could.
The courier cash cycle is a working capital problem
Here is the trap that kills growing stores specifically.
You pay your supplier for stock. You pay for advertising. The customer pays cash at the door. The courier holds that cash and remits it to you on their own cycle, days or weeks later.
So your cash goes out immediately and comes back late, and the faster you grow, the bigger the gap gets. Profitable stores run out of money this way routinely, and the owner cannot understand it because the profit and loss statement looks fine.
Three defences:
- Know your courier's remittance cycle exactly and model it. Negotiate it once you have volume.
- Reconcile courier remittances to orders every week. Discrepancies are common, and unreconciled after a month they are unrecoverable.
- Run a rolling cash forecast. This is precisely what the thirteen-week cash flow forecast exists for, and a COD store is the clearest case for it.
Marketplace, own store, or social
| Channel | Gets you | Costs you |
|---|---|---|
| Marketplace | Immediate traffic, trust, established logistics | Commission, price competition, no customer relationship, no data, and a platform that can change the rules |
| Own store | Margin, customer data, brand, a retargetable audience | You must generate every visit yourself |
| Social selling | The cheapest start, conversational trust, strong for discovery | Manual, hard to scale, fragile if an account is lost |
The sequence we recommend for most new sellers: prove demand on a marketplace or through social, build the own-store in parallel, then shift the repeat customers across. The marketplace is customer acquisition. The own store is where the margin and the asset live. Treating either as the whole strategy leaves money on the table.
If you are building the store properly, that is ecommerce store setup, and the sector view is under ecommerce.
Social commerce, done as a funnel rather than a vibe
Short video is where Pakistani product discovery increasingly happens, and most brands treat it as content rather than as a funnel. The version that works:
- Discovery content that shows the product in use, not a catalogue post.
- A single, obvious next step, a link or a comment keyword that starts a conversation.
- A fast, human reply, within minutes, not hours. Response time is the conversion variable.
- A checkout that does not lose them, which usually means the order can be completed inside the chat.
- The customer record ends up in one system, not scattered across three inboxes. A CRM matters more here than people expect.
Paid amplification works once the organic version converts, not before. That is the order: social media marketing first, then PPC behind a proven funnel.
A ninety-day build
| Weeks | Focus |
|---|---|
| 1 to 2 | Catalogue, pricing with real landed cost, address-quality checkout, prepaid incentive live |
| 3 to 4 | Courier selected, remittance cycle documented, confirmation process running for every order |
| 5 to 8 | One acquisition channel worked properly. Return rate and contribution per order tracked weekly |
| 9 to 12 | Repeat purchase push to existing customers, blocklist in use, weekly reconciliation routine, honest review of contribution margin |
Frequently asked questions
Why do Pakistani online stores have such high return rates?
Because cash on delivery gives the customer a free option and asks for no commitment at purchase. The order costs them nothing until the courier arrives, so impulse orders, duplicate orders and simple absence all convert into refusals that you pay for.
Should I sell on a marketplace or build my own website?
Both, in sequence. A marketplace gives you demand and trust quickly. Your own store gives you margin, customer data and an asset you own. Use the first to find customers and the second to keep them.
How do I reduce cash on delivery refusals?
Confirm every order before dispatch, collect a partial advance on high-risk orders, capture structured addresses with a landmark, message the customer on the delivery day, and require prepayment from repeat refusers.
How much stock should I hold when I start?
Less than feels comfortable. Cash tied up in stock is the other half of the working capital squeeze that COD creates. Start narrow, prove which items actually sell through, and widen the range from sales data rather than optimism.
Is social commerce better than a website in Pakistan?
It is better at discovery and worse at scale. Sell through social by all means, and make sure the customer ends up in your own records and, over time, on your own store. An audience on a platform is rented; a customer list is owned.
Where to go next
If the store needs building or fixing, ecommerce store setup covers the build, the payments and the operational wiring. If you also sell in person, retail shop setup keeps the two sides on one set of numbers. If you are not sure whether your unit economics work at all, send us your last hundred orders and we will tell you.
