How do you optimize cash on delivery and cut refusals?

Why do customers refuse COD orders?

Almost every refusal traces to one of five causes. Impulse cooling: the order was emotional, and three days later the emotion is gone. Zero commitment: nothing was paid, so refusing costs the customer nothing and no guilt. Slow or unpredictable delivery: refusal probability climbs steeply after day three, and every failed first attempt makes it worse. Surprise totals: shipping or COD fees the customer did not register at checkout. And absence: the courier came, the customer was at work, the second attempt never really happened.

Notice what is not on the list: 'bad customers'. A small fraction are serial refusers you should block — but the majority of refusals are process failures, which is good news, because process is fixable.

What does a WhatsApp order confirmation flow look like?

Confirmation is the single highest-impact COD fix: it converts a zero-commitment order into a verbal promise, catches wrong numbers and addresses before you pay for shipping, and filters the orders that were never serious. The flow that works:

• Within the first hour of the order: 'Hi [name], your order from [store] is confirmed for prep: [item], total [X] including delivery, arriving [day]. Reply 1 to confirm or 2 to change anything.' Total must include every fee — surprises at the door are refusals.
• No reply after a few hours: one gentle nudge, then a phone call for high-value orders. Orders that never confirm ship last or not at all — track how this rule alone changes your refusal rate.
• At handover to the courier: 'Your order is with [courier], tracking [link], expected [day].' Named courier plus a real date measurably reduces at-door refusals.
• Delivery day: 'Arriving today between [window]. Please keep [amount] ready — or pay now by card and skip the cash: [link].' That last line quietly converts COD orders to prepaid even after checkout.

How do you move customers from COD to prepaid?

You will not eliminate COD in the Gulf and should not try — for a first-time customer it is a legitimate trust bridge. The goal is shifting the persuadable middle. Build an incentive ladder and climb it gradually: free or discounted shipping for prepaid orders; a small prepaid discount (even a token amount changes behavior); a visible-but-modest COD handling fee framed as the courier's cash-collection cost; and prepaid-only for sale items and heavy discounts, where refusal risk is highest and margin thinnest.

Two Gulf-specific accelerators: local payment methods and BNPL. Offering Mada and Apple Pay in Saudi Arabia removes most card friction — Apple Pay especially converts customers who would never type a card number. And Tabby or Tamara installments give the customer COD-like safety (pay later) while you get prepaid-like commitment. For many stores, BNPL is the single biggest COD-to-prepaid shifter.

Which delivery-side fixes cut refusals the most?

Speed first: ship within 24 hours of confirmation. The refusal curve versus delivery time is the most consistent pattern in COD — same-city next-day delivery refuses at a fraction of the rate of five-day delivery. If your courier cannot do 48 hours to major cities, that courier choice is costing you more than their invoice shows.

Then the mechanics: choose couriers that send WhatsApp updates and call before arrival (a customer who chose a delivery window rarely refuses at it); validate addresses at checkout — in the Gulf, a required 'nearest landmark' field plus a shared location pin saves more packages than any script; and script the second attempt properly, offering an alternative window or pickup point instead of silently returning the parcel. Ask your courier for refusal rate by driver and area — the spread will surprise you, and it is negotiating leverage.

When should you block COD for a customer or product?

COD should be a privilege with rules, not a universal right. Three rules cover most cases. Repeat refusers: two refused orders means prepaid-only for that phone number — keep a simple blocklist; serial refusers are a tiny share of customers but a heavy share of losses. High-ticket thresholds: above a value you choose (for many stores, several hundred SAR), require prepayment or at least a small card deposit — one refused premium item can erase the margin of ten delivered ones. New-customer risk: first-time, high-value, un-confirmed on WhatsApp, rush sale period — any two of those together justify a verification call before shipping.

Announce none of this loudly. The checkout simply shows the payment options each customer has earned.

How do you measure COD performance properly?

The headline metric is refusal rate: refused deliveries divided by shipped COD orders. Track it weekly, and never as one blended number — split it by product, city, courier and campaign. A store-wide 15% often decomposes into a well-behaved 7% core plus one product, one remote city or one traffic source refusing at triple the rate. That decomposition is the to-do list.

Then make refusals visible in your marketing math: a refused order should subtract its shipping-both-ways and handling cost from the campaign that generated it. Platform ROAS counts the sale the moment it is placed; your bank account counts it when cash is collected. Stores that re-rank campaigns on delivered revenue instead of placed revenue routinely discover their 'best' campaign was quietly their worst — usually the one leaning hardest on impulse.