Why cash on delivery is popular in some markets and not others

Sellers in Jakarta and Ho Chi Minh City and Quezon City get told the same thing by every consultant: educate your buyers, push them to prepay, cash on delivery is a legacy habit that will fade.

It hasn't faded, and the advice misreads why it exists.

Cash on delivery is popular because it is the only guarantee the buyer can enforce by themselves. Not because people lack cards, and not because they distrust technology. In a market where a buyer who gets scammed has no realistic path to their money back, refusing to pay until the box is in their hand is the entire consumer protection system. That's why digital wallets arriving didn't kill it. And it's why it costs you so much: GoKwik's data across more than 180 million Indian shoppers put COD orders near a 26% return-to-origin rate against under 2% on prepaid, with fashion and footwear touching 40% (their figures, read August 2026). You're not offering a payment option. You're underwriting a loan and paying the interest yourself.

COD is a guarantee, not a payment method

Think about what the buyer is actually buying with it.

They're not buying convenience. Handing a rider a Rp 95,000 note is not more convenient than tapping a wallet they already have open. They're buying certainty that they will not pay for nothing.

That's a recourse question, and recourse is what varies between markets. Where card chargebacks reliably work, where a marketplace escrow actually pays out, where a small claims process exists and is usable by an ordinary person, prepaying is cheap because the risk is somebody else's. Where none of that is dependable, the buyer keeps the only lever they have, which is their own money, until the moment the goods exist in their hands.

Cash on delivery is what buyers use instead of a legal system.

Once you see it that way, most of the confusing behaviour makes sense. A buyer with three payment apps and a credit card still picks COD from a seller they found yesterday, and prepays the same amount to a seller they've bought from four times. Their card availability never changed. Their exposure did.

What the buyer is actually afraid of

It's specific, and it isn't fraud in the card sense.

Getting nothing. Money leaves, parcel never arrives, seller stops replying. The account is gone in a week.

Getting something else. The bag is a different bag. The fabric is thinner. The phone case fits a different model. Counterfeit and substitution are lived experience in a lot of these markets, not an abstract worry.

Getting nothing back. Even honest disputes stall. Returning an item to a seller who doesn't want it returned is a fortnight of messages and a courier fee.

Notice that all three are about the seller being unknown, not about money being digital. Which is why the thing that actually reduces COD is not payment education. It's you becoming known.

Why wallets didn't kill it

Indonesia has QR standards accepted at warungs. The Philippines has GCash on essentially every phone. Vietnam has bank QR transfers that clear in seconds and cost nothing. India rebuilt person-to-person payments from scratch.

COD is still enormous in all of them.

Because the payment instrument was never the constraint. Swapping cash for a wallet at the door doesn't change who is exposed; it just means the rider watches a QR scan instead of counting notes. In fact the most common modern form of "cash on delivery" in several of these markets is a digital transfer made at the doorstep, with the rider standing there. Same guarantee, no cash. The name is out of date and the mechanism isn't.

That's a useful thing to understand, because it tells you which lever to pull. If your buyers COD because they don't trust you yet, adding another wallet does nothing. If they COD because handing over cash is genuinely easier for them, a doorstep QR helps.

The three bills you get for offering it

They arrive separately, which is why sellers underestimate the total.

The bill	How it lands
Refused and returned parcels	Freight out, freight back, and an item that comes home creased, opened or missing a tag
The float	The courier holds your money for days after delivery. Your working capital is in a van
Handling	A percentage or flat fee on the collected amount, taken before you're paid

Then there's the fourth one nobody bills you for: your time. Chasing courier remittances, reconciling which of last week's forty deliveries were actually collected, and arguing about a parcel marked delivered that the buyer says never arrived.

The arithmetic that matters is refusals against margin, not against price. Sell a Rp 95,000 item with Rp 40,000 of cost in it, with a round trip costing Rp 22,000, and one refusal in eight eats about a third of the profit on the other seven. Sell a Rp 450,000 item and you can absorb it. Anyone selling low-value items on COD with thin margins is running a business that only looks profitable because the returns arrive a month after the sales.

Cash on delivery for small sellers goes through refusal handling, change, and when to switch to a deposit instead.

The refusal doesn't happen at the door

This is the part experience teaches and articles don't.

When a buyer refuses a parcel, the decision was almost never made at the door. It was made on day four, when they saw a similar thing cheaper, or the money got spent on something else, or the impulse simply passed. The parcel then arrives on day nine into a mood that no longer wants it.

Which means your single biggest lever on refusal rate is not your payment terms. It's transit time. Every day between the order and the doorstep is another day for the buyer's mind to change, and cutting three days off delivery does more for your refusal rate than any amount of confirmation messaging.

Go and look at your own returned parcels from the last three months. Line them up against how long each one took to deliver. Most sellers who do this find a threshold, somewhere around day five to day seven, past which refusals climb sharply. Find yours, then either beat it or stop promising past it.

The second lever is contact. A parcel out for delivery to a buyer who doesn't answer an unknown number comes back regardless of how much they wanted it. One message on dispatch day, telling them to expect a call, is the cheapest return-reduction available.

Worked example: Dewi, prayer sets, Bandung

Dewi sells mukena and prayer sets from her home in Bandung. Rp 95,000 for the standard set, Rp 185,000 for the embroidered one. Around 120 orders a month, sold through Instagram and a WhatsApp broadcast list, shipped across Java and out to Sumatra and Kalimantan.

She used to offer COD on everything. Her numbers were unpleasant once she wrote them down: roughly one parcel in six coming back, concentrated on the Sumatra and Kalimantan orders where transit ran seven to eleven days. On a Rp 95,000 set with about Rp 22,000 of two-way freight, that was quietly eating most of a month.

What she changed, in this order:

1. Geography. COD on Java only. Everywhere else, prepaid, with the reason stated plainly on her page: "shipping outside Java takes 7 to 11 days, so those orders are prepaid."
2. A dispatch message. One line on the day the parcel leaves, with the tracking number and a note to answer an unknown number.
3. A small prepaid discount. Rp 5,000 off for paying in advance. Not a punishment for COD, a nudge toward prepaid, which reads completely differently to a buyer.
4. A confirmation reply. Any COD order over Rp 300,000 gets a "confirm this is still what you want" message before dispatch. About one in twelve says no, and every one of those is a parcel she didn't pay to send twice.

Returns went from about one in six to about one in fourteen. Prepaid share went from roughly a quarter of orders to just over half, and almost all of that shift came from repeat buyers, not from persuading strangers.

That last detail is the whole lesson. Nobody talked a first-time buyer out of COD. Buyers moved to prepaid after they'd bought once and it went fine.

What actually moves people to prepaid

In order of effect:

Having bought from you before. Nothing else comes close. The second order from a happy customer is where prepaid becomes available to you.

Visible history. Customer photos, dated posts going back months, replies from real people. A buyer deciding whether to prepay is looking for evidence you'll still exist next week. Building trust before money moves is the practical version of this.

A stated, believable return policy. Not a generous one. A specific one. "If it's wrong, message me within 3 days with a photo and I'll send a replacement or refund" beats a paragraph of legal-sounding text because it tells them what will actually happen.

A small, honest incentive. A discount, free shipping, faster dispatch. Small is fine. The point is to give a reason, not to buy the behaviour.

What doesn't work: telling buyers COD is expensive for you. That's your problem, not theirs, and framing it as their inconvenience reads badly.

Where Sailo fits, and what it doesn't do

Sailo has cash on delivery as a payment rail, and it takes nothing on it, because on COD it never touches the money. What you get is the order recorded properly, with the item, the options, the address and the total, instead of forty messages you'll have to scroll back through when the rider calls.

The COD rail itself is deliberately thin: one free-text delivery notes field. Use it for the things that actually cause refusals. Which areas you serve, how many days it really takes, whether you can break a large note, and what you'd like the buyer to do on dispatch day.

The honest limits:

Manual rails mean you confirm. Sailo does not know whether the rider collected. It does not chase your courier's remittance. The order sits where you left it until you move it, and that reconciliation is your job, every day, forever. Rider handovers and cash reconciliation is how to do it without losing money in the gaps.

There's no mobile wallet rail. No GCash, no QRIS integration, no wallet of any kind. The full list is card, WhatsApp, Telegram, Instagram, email, phone, bank transfer and cash on delivery. You can put a wallet number into the bank transfer instructions field and confirm it yourself in your own app, and plenty of sellers do, but that's a workaround you're running, not a feature you're using.

What to do next

Pull your last three months of returned parcels and put two columns next to each one: days in transit, and whether the buyer had ordered from you before. That table will tell you more about your own business than this article can.

If transit time is the pattern, change your delivery promise or your courier before you change anything about payment. If first-time buyers are the pattern, your problem is trust and the fix is proof rather than policy.

Then pick one geography where COD is costing you the most and try prepaid-only there for a month, with the reason written plainly on the page. If your orders from that region don't drop, you've just found free money.

For the wider comparison of what each payment rail costs and what it leaves on your desk afterwards, how to take payment as a small seller covers all of them. And if you sell into a market where COD is the default expectation, how to sell online in the Philippines as a beginner works through the same problem from the seller's side of it.