Cash on delivery: when it helps sales and when it eats the margin

Turn COD off and orders drop. Turn it on and the courier keeps bringing parcels back. Cash on delivery helps sales when the buyer has never heard of you and will not risk paying first. It eats the margin when parcels are refused, because you pay shipping both ways and lose the ad money that won the order. Offer it where your own refusal share stays below the share you can afford, confirm every COD order before dispatch, and switch it off for products a refusal ruins.

Why do buyers still pick COD?

A buyer choosing COD is not always telling you they have no UPI. Often they are saying they do not trust you yet. They have seen the reel and liked the product, and have no idea whether the parcel will turn up, or whether a refund would ever arrive if it is wrong. COD moves that risk onto you. For a brand with no reviews and no name, that is a fair price for a first order.

The second order is different. The buyer has the parcel, knows you deliver, and paying ahead is easier than keeping exact change ready at the door. If repeat buyers still choose COD, ask a few of them why.

What does a refused parcel actually cost?

Owners count the courier charge and stop. The real cost of a refused COD parcel, a return to origin or RTO, has more lines:

• Forward shipping, paid even though nothing was sold
• The return leg, which courier contracts commonly charge as well; check yours
• Any COD handling charge your courier or shipping service adds
• Packing material, labels and the time it took to pack
• Stock out of circulation while it travels, sometimes your best-selling size
• Damage: a crushed box, soiled fabric, a broken seal that makes it unsellable as new
• And the ad money spent to win that order, which never comes back

That last line turns a small loss into a large one.

Work out the refusal share you can afford

Fill this in for one product at a time, with figures from your own shipping dashboard and ad account.

Illustrative figures in rupees for one product. Use your own.

A  Price the buyer pays                 699
B  Product cost                         260
C  Forward shipping                      70
D  COD charge                            30
E  Packing                               15
F  Return leg if refused                 70
G  Damage on refused parcels, averaged   25
H  Ad spend per order                   220

Margin on a delivered order    M = A - B - C - D - E   = 324
Loss on a refused parcel       L = C + E + F + G       = 180
Refusal share you can afford     = (M - H) / (M + L)   = 104 / 504, about 1 in 5

At one refused parcel in five, this product stops making money on COD. Now look at line H. Halve the ad spend per order and the share you can afford roughly doubles. A higher-priced product with the same shipping can also afford more refusals, because shipping is a smaller part of what it earns. So one COD rule for the whole catalogue will be wrong somewhere.

Then pull your actual refusal share for COD orders, by product and by price band. Where it sits below the affordable share, COD is paying for itself. Where it sits above, every COD order on that product loses money before anyone notices.

Where to keep COD on, and where to switch it off

Keep it for first orders on standard products that can go straight back on the shelf, wherever your actual refusal share is under the line.

Personalised and made-to-order pieces are the clearest case for switching it off. A refused name-engraved pen or a blouse stitched to measure cannot be sold to anyone else. Make those prepaid, or take a UPI advance that covers the making and collect the rest on delivery.

Low-priced items where the worksheet says no come next. Put a small prepaid discount on the product page, where the buyer sees it before deciding, not only at checkout.

Before you block any pin code, check how long deliveries there take. A parcel that arrives on day nine meets a buyer who has already bought elsewhere, and the fix for that is the courier, not the buyer.

Do not switch COD off across the whole store on the strength of one bad fortnight. Change it on one category or price band, keep everything else as it was, and compare orders and refusals with the weeks before.

For example: juttis from Jaipur

For example, picture an invented Jaipur seller of hand-embroidered juttis on its own store, with a custom range where buyers pick the thread colour to match a wedding outfit. The figures are made up.

Instagram ads with "COD available" in the caption brought 90 late-night orders in a week before the wedding season. The seller shipped every one, standard and custom, without a word to the buyer. Nineteen came back refused. The standard pairs came back and went back on the shelf, minus two lots of shipping. The custom pairs came back and could not be sold at all, because nobody else wanted mustard thread on a size 4. Eleven custom pairs, about ₹2,400 each in making cost, were written off.

The fix was a rule per range. Custom juttis became prepaid, or COD with a UPI advance that covers the embroidery. Standard pairs kept COD, with a confirmation message before dispatch. And "COD available" came out of the ad captions. The partial result: refusals fell to about one in ten, not zero, and some buyers who wanted COD for custom pairs went elsewhere.

Confirm before you ship

A COD order nobody has confirmed is a guess. A short WhatsApp message with the item, the amount to keep ready and a one-tap cancel catches the forgotten midnight order before you pay the courier. Our guide to confirming a COD order on WhatsApp has the message and the call script.

What to do next

Fill in the worksheet for your three best sellers using your own dashboard. If one shows COD losing money, change that one first and leave the rest alone. Which payment options suit each product category is covered in the e-commerce sales and marketing section, and lead response and follow-up is the service for deciding who phones the doubtful orders. To talk it through, tell us what you need.