A refused COD order costs you 4–7 KD, not nothing
Most owners book a failed COD delivery as zero. It isn't. Count it properly: the courier still charges you for the attempt, usually 1.5–2.5 KD in Kuwait, and often close to the same again for the return leg. Add packaging you can't reuse, the pick and pack labour, and the hour your team spends chasing the customer on WhatsApp. Most stores land between 4 and 7 KD of real cost per refused order, before you count the stock that sat in a van for three days instead of being sold to someone else.
Now put that against your margin. If you sell a 12 KD abaya at 5 KD gross margin and one order in six comes back, those five delivered orders earn 25 KD and the refusal eats 5 to 7 of it. You just lost a quarter of the profit on that batch. This is why COD-heavy stores in Kuwait can grow revenue every single month and the owner still feels broke. The number that matters is not orders shipped. It is orders delivered and collected.
Then there is the cash itself. Your driver or courier holds your money for seven to thirty days depending on the settlement cycle, and you reconcile it by hand against a spreadsheet. Short counts happen. Notes go missing on a Friday. Small stores routinely find 20–60 KD a month they cannot trace and quietly write it off. If you are doing 300 COD orders a month you need one person who does nothing but match cash to order numbers, and that person is a real cost sitting on top of the delivery fees.
Ready to start your Shopify store?
Start a free trial and try the platform for yourself.
COD hides a conversion problem, it does not fix it
Refusal rates tell you what your marketing is really doing. On repeat customers and warm traffic, Kuwait stores usually see 3–6% of COD orders refused. On cold Snapchat or TikTok traffic with an impulse product, 15–25% is normal and 30% is not rare. Same store, same courier, same driver. The difference is whether the buyer actually decided to own the thing or just tapped a button at 1am. COD lets that undecided tap become an order in your dashboard, so your reported conversion rate looks healthy while your delivered conversion rate is half of it.
That is the trap. COD does not create demand. It postpones the decision to the doorstep, where you have no salesmanship left and a driver in a hurry. Every store we have cleaned up showed the same pattern: the products with the highest refusal rate were the ones with the weakest product pages — no size chart, no real photos, no clear return policy, no delivery time stated. Fix the page and refusals drop before you touch payment at all. If refusals are above 15%, treat it as a merchandising problem first and a payment problem second.
It is also worth checking that prepaid is genuinely easy on your side. Plenty of Kuwaiti stores advertise online payment on a checkout that breaks on iPhone, redirects twice through KNET, and demands a full address before it shows the total. Customers pick cash because the alternative is painful, not because they distrust you. Moving to a checkout that handles KNET, cards and Tabby cleanly shifts five to fifteen points of volume to prepaid on its own, with no discount offered and no argument with the customer.
Confirm it, charge for it, and cap it
Confirm every COD order before it leaves the warehouse. One WhatsApp message within the hour: name, item, total, address, and a yes. No reply after two attempts across a day means the order does not go on the van. This single habit takes most stores from a 15% refusal rate down to 6–8%, and it costs you a part-timer for two hours a day. Send it from the store's business number, keep it short, and never make it sound like you are checking up on them. "Confirming your order before we ship today" is enough.
Then price COD honestly. A 1 KD COD fee, waived above a 25 KD basket, is completely normal in Kuwait now and almost nobody abandons over it. Pair it with a real prepaid incentive: 5–10% off, or free delivery on prepaid only. You are not punishing cash customers, you are letting the customers who cost you less pay less. Most stores see prepaid share move from around 20% to 45–60% within three months of running both levers together, and refusals fall with it because a prepaid order almost never comes back.
Finally, cap COD instead of killing it. Set a ceiling — say 40 KD — above which the order must be prepaid, because a refused 90 KD order hurts far more than a refused 12 KD one. Block COD on first orders from brand-new numbers if you get a wave of fake orders, then open it back up once that customer has taken delivery once. Keep a quiet blocklist of numbers that refused twice. That gives you most of the benefit of removing COD without losing the 25–40% of order volume that switching it off cold would cost you.