COD Return Rates in Pakistan: Measure Yours, Then Cut It

Nobody can tell you what the COD return rate in Pakistan is. There is no audited national figure, and the numbers thrown around in reseller groups belong to someone else's products, someone else's ads and someone else's courier. What you can know exactly is your rate, and that is the only one you can act on.

Your COD return rate — resellers call it RTO, return to origin — is the share of dispatched parcels that come back undelivered. The formula is:

RTO rate = parcels returned ÷ parcels dispatched × 100, measured over a fixed window.

Everything useful follows from that one number: what a failed parcel costs you, which products and which cities produce your returns, and which fix below is worth doing first.

What actually counts as a return

Four different things get called "a return", and lumping them together is why most resellers cannot fix theirs. Separate them:

• Cancelled before dispatch. The order never turned into a parcel — the buyer did not confirm, or confirmed and backed out. This is not RTO. It is a leak in your confirmation step, and it is the cheapest one to fix because no delivery charge was spent.
• Undelivered. The courier could not complete the delivery: phone switched off, wrong or vague address, nobody at home after the attempts allowed. The parcel goes back without the buyer ever seeing it.
• Refused at the door. The buyer saw the parcel and declined to pay. This one is the most informative, because something between your ad and their doorstep did not match.
• Returned after delivery. They paid, then sent it back inside the return window. This costs you differently — the money arrived and then left.

Your RTO rate is the middle two. Track the other two beside it, in the same sheet, because a fix in one often moves another.

How to calculate your rate properly

The formula is easy. Applying it to a moving business is where people get a number that lies to them.

Count by dispatch cohort, not by calendar month

A parcel dispatched on the 28th will not resolve until well into the following month. If you divide this month's returns by this month's dispatches, you are dividing one group of parcels by a different group, and the result jumps around for reasons that have nothing to do with your business. Instead, tag each parcel with its dispatch date and wait until that batch is fully resolved — delivered, refused or returned — before you score it. A batch that is still in transit is not a number yet.

Segment it, or it tells you nothing

A single blended rate tells you that you have a problem. It never tells you where. Break the same numbers down four ways:

• By product. Almost always the biggest spread. A handful of items usually generate most of the failures.
• By city or region. Delivery reliability is not uniform, and some routes need a different confirmation habit than others.
• By order source. Facebook ad, Instagram DM, WhatsApp status, TikTok, a repeat customer — these behave very differently.
• By order value. The higher the cash the buyer has to produce at the door, the more the door becomes a decision point.

You do not need software for this. A spreadsheet with one row per parcel and columns for product, city, source, value, dispatch date and outcome will out-diagnose any dashboard you could buy.

Track the whole funnel, not just the last step

Orders placed → confirmed → dispatched → delivered → paid → returned. Write down the count at every stage. A reseller with a low RTO rate and a terrible confirmation rate is not doing better than one with the reverse — they are just failing earlier, and more cheaply.

What one refused parcel actually costs

The instinct is to treat a return as a wash: no sale, no loss. It is not a wash. A failed COD parcel charges you on both legs plus everything you spent to create the order.

• The forward delivery charge, which most courier arrangements still bill on a parcel that goes out and comes back.
• The return leg, billed separately in most rate sheets.
• Packaging, which rarely survives a round trip in resaleable condition.
• What the order cost to win — ad spend, or the hours you spent on DMs and calls.
• Stock sitting in a van instead of being available for a customer who would have paid.
• Your time, twice: once dispatching, once receiving and restocking.

Work out your own figure once. Call the total cost of a failed parcel F, and your margin on a delivered order M. Then the number of successful deliveries it takes to pay for one failure is F ÷ M. Fill in your courier's rate sheet and your actual margin and you will have a number that changes how you think about the whole operation.

The arithmetic is unforgiving at the edges. Take a batch of four dispatches at a 25% return rate: three deliver, one comes back. If that one failure costs the margin of three deliveries, the batch nets you roughly nothing. The same batch at a 10% rate is a working business. That gap is not a rounding difference — it is the difference between a reseller who keeps going and one who quietly stops.

This is also why a higher average order value is worth chasing before a higher order count: the cost of a failure is largely fixed, so a bigger delivered order absorbs it better. Product bundling to raise your average order value covers the practical way to do that. If you are comparing where to sell, Daraz vs Markaz for resellers walks through how fees and COD terms change the same maths.

Where your returns actually come from

Six causes cover nearly all of it. Your segmented sheet will tell you which ones are yours.

The order was never real

One-tap order forms, giveaway-style ad creative and open comment sections produce orders nobody intended to place. Wrong numbers, joke entries, duplicate submissions from a slow page. These never had a chance of being delivered and they are pure cost if you dispatch them.

They forgot, or changed their mind

An impulse order placed at midnight looks different in daylight three days later. The longer the gap between the order and the doorstep, the more of these you get. Speed is a returns lever, not only a service one.

The product did not match the listing

The photo was better lit than the product, the size ran small, the colour read differently on a phone screen, the material felt cheaper in hand. The buyer is refusing the gap between the two, not the product.

The total at the door was a surprise

If the delivery charge was never stated, the courier's number is a bigger figure than the one the buyer agreed to. People refuse surprises even when the surprise is small.

Nobody could reach the buyer

A number that goes to voicemail, an address that stops at a sector name, a rider who arrives during working hours at a house nobody is in. Most of these are fixable at order time, not at delivery time.

It arrived too late

The occasion passed, or they bought it locally in the meantime. A parcel that arrives after the buyer has moved on is a refusal waiting to happen.

The levers, in the order they pay off

Work down this list. The early items cost nothing and move the number most.

Confirm every order before it is dispatched

One call or one WhatsApp message that names the product, the total including delivery, and the expected delivery day. An order that cannot be confirmed after two attempts should not become a parcel. This single habit removes most of the junk-order category before it costs you anything. Our shorter companion guide, how to reduce COD returns (RTO) in Pakistan, has the confirmation script and the wording that works.

Fix the listing before you blame the courier

If one product's return rate is far above your average, the problem is upstream of delivery. Use real photos from more than one angle, state measurements rather than sizes alone, and describe the material honestly. A listing that undersells slightly and over-delivers at the door produces almost no refusals.

Quote the door total, always

Product price plus delivery charge, stated as one number the buyer agrees to before dispatch. Never let the rider be the one who introduces a new figure.

Set the delivery expectation, then meet it

Tell the buyer the realistic window and keep them posted when the parcel moves. A buyer who knows a parcel is arriving Thursday plans to be there on Thursday. A buyer expecting it "soon" is a coin flip.

Make the address collectable

Capture a nearest landmark, a second phone number and a preferred time window at order time. Riders return parcels they cannot deliver, and a surprising share of those are addresses that were never quite complete.

Choose the courier for its reattempts, not its sticker rate

The cheapest per-parcel rate is not the cheapest per delivered parcel. How many delivery attempts are made, whether the rider calls before arriving, how quickly a return is brought back to you, how the COD payment is remitted — those terms move your real cost far more than a small difference in the forward charge. Compare couriers on the delivered rate you measured, not on the price list.

Cut the products that return

Once you have per-product rates, the decision makes itself. Some items are not worth selling on cash on delivery at all, however good the margin looks on the ones that do land.

Fix the ad, not just the order

Return rate is partly bought at the top of the funnel. Creative that overstates, prices that only appear after the click, and lead forms that collect a tap instead of an intention all produce orders that fail later. Filtering harder at the ad stage looks like fewer orders and is usually more delivered parcels.

The categories that come back most

Patterns worth knowing before you build a catalogue:

• Clothing and shoes — sizing is the single most common reason a parcel is refused or returned after delivery. Publish measurements, not just S/M/L.
• Colour-critical items. Anything a buyer pictures in a specific shade will disappoint some share of buyers, because screens differ.
• Fragile goods. A damaged arrival is a guaranteed refusal and a total loss on the item.
• High-ticket impulse buys. The bigger the cash at the door, the more second thoughts.
• Electronics. Buyers expect to test at the door and expect a warranty story you may not be able to give them.

None of these are unsellable. They need a tighter confirmation call, a more honest listing and a higher margin to absorb the failures they will produce.

When a parcel does come back

Handle returns as a process, not an annoyance.

• Open and inspect it the day it arrives. Damage found a week later is nobody's fault but yours.
• Restock it fast — it is cash sitting on a shelf until it is listed again.
• Log the reason in the same sheet, in the buyer's words where you have them. The reason column is what makes next month's rate lower.
• Note repeat refusers. A short list of numbers worth confirming twice is worth keeping.
• Repack properly before it goes out again. Packaging and branding for small online sellers in Pakistan covers what a COD parcel has to survive.

How the model you sell on changes the maths

Everything above assumes the returns risk is yours to carry. How much of it you carry depends on how you are set up.

If you buy stock upfront and ship it yourself, every failure is fully yours: the stock, both delivery legs, the packaging and the cash tied up in inventory that did not move. That is the traditional model, and it is why return rates matter so much to it.

The reseller model on the Markaz dropshipping portal changes the shape of that risk. You do not buy stock in advance — there is no minimum order quantity, so you order a unit after a customer has bought it. Deliveries run inside Pakistan in roughly 3–5 days on cash on delivery, and the platform handles the courier leg and the COD collection, so you are not negotiating rate sheets or chasing remittances. Markaz also runs a return window and buyer protection, so a return is handled within the platform's policy rather than leaving you holding stock you cannot sell.

That does not make returns free. A refused parcel is still a lost sale, still spent ad money, and still a signal that something in your listing or your confirmation habit needs work. What it changes is the worst case: you are not sitting on a carton of an item that turned out to have a 40% refusal rate. If you are setting up on this model, the Markaz reseller programme is where the no-MOQ catalogue and COD terms are laid out, and you can see what moves on cash on delivery in the Markaz deals. Start with COD dropshipping in Pakistan if the model itself is new to you.

A 20-minute weekly review

Once a week, with your sheet open:

• Score last week's fully-resolved dispatch batch. One number.
• Sort by product. Name the worst two.
• Sort by source. Is one ad or one channel producing your failures?
• Read the reason column. Pick the one cause that appears most.
• Change exactly one thing this week — a listing, a confirmation script, a product you stop selling, an ad you turn off.
• Compare next week's batch to this one.

One change per week, measured, beats ten changes at once that you cannot attribute. Six weeks of this will move your rate further than any single tactic.

Worth reading next: the best dropshipping apps with cash on delivery in Pakistan compares how different platforms handle the COD leg, and how much money you need to start dropshipping in Pakistan shows where returns fit into the starting budget.

Frequently asked questions

What is a good COD return rate in Pakistan?

There is no published national benchmark, and any single figure you are quoted comes from one seller's catalogue, ads and courier. The useful target is your own trend: measure your rate this month, change one thing, and measure the next fully-resolved batch. A rate that is falling month over month is the only benchmark that means anything.

How do I calculate my RTO rate?

Divide the parcels returned by the parcels dispatched in the same batch, then multiply by 100. Score a batch only once every parcel in it has resolved, and tag each parcel with product, city, order source and value so you can break the number down.

Does a returned COD parcel still cost me money?

Yes. Most courier arrangements bill the forward leg even when the parcel comes back, and the return leg is usually charged as well. Add the packaging, whatever you spent to win the order, and the days the stock spent in transit instead of on sale.

What is the fastest way to lower my return rate?

Confirm every order before dispatch, and quote the door total including delivery. Those two habits cost nothing, need no tools, and remove the two largest categories of failure — orders that were never real, and buyers surprised by the amount the rider asks for.

Which products have the highest COD return rates?

Typically anything size-dependent such as clothing and shoes, colour-critical items, fragile goods, and high-value orders where the buyer must produce a lot of cash at the door. Measure your own per-product rates rather than assuming — your photos and your confirmation call change the outcome as much as the category does.

Can I avoid the returns risk entirely?

No, but you can decide how much of it you carry. Buying stock upfront puts the whole risk on you. Selling on a no-MOQ platform such as Markaz means you order a unit only after a customer buys it, the platform handles the courier and COD collection, and a return is dealt with under the platform's policy instead of leaving you with unsold stock.