Cash-on-Delivery Confirmation Rate: What Moves It and How to Measure It

In a cash-on-delivery business, an order on the screen is not yet an order. Someone has to answer the phone or the message, say yes, and still be there when the courier arrives. The share of orders that survive the first of those steps is the confirmation rate, and many owners quote it without being able to say exactly what it counts.

The decision in front of you is simple: where in the chain do you lose orders, and is that loss worth fixing before you spend more on ads?

Why one number misleads

"Our confirmation rate is 60%" can mean at least three different things. It may be confirmed orders divided by all orders placed. It may be confirmed divided by orders you managed to reach. Or it may be confirmed orders that were later delivered.

These give different answers on the same data, and each points to a different fix. A low rate caused by unreachable customers is a contact problem. A low rate among customers you did reach is an offer, expectation or trust problem. Mixing them sends you to fix the wrong thing.

A four-step model with a measure point at each step

Treat confirmation as a funnel with one count and one rate per step. Define each count in writing, once, and keep it fixed.

1. 1

   Order placed

   Count every order that reaches your system

   100
2. 2

   Customer reached

   A real conversation or a reply, not just an attempt

   80
3. 3

   Order confirmed

   The customer explicitly agrees to receive it

   60
4. 4

   Delivered and paid

   Confirmed orders that complete

   46

Reach rate 80%, confirmation among reached 75%, overall 60%, delivered of placed 46%.

Four rates come out of it:

1. Reach rate: reached ÷ placed.
2. Confirmation among reached: confirmed ÷ reached.
3. Overall confirmation rate: confirmed ÷ placed.
4. Completion rate: delivered ÷ confirmed.

The second and fourth are the ones that explain behaviour. The third is the one to report.

Reach rate

80%

Confirmation among reached

75%

Overall confirmation

60%

Completion of confirmed

77%

46 ÷ 60 confirmed is about 77%.

What tends to move each step

These are hypotheses to test, not guaranteed levers.

• Reach: speed of first contact, the channel used, whether the phone number was captured correctly at checkout, and the hours when you try.
• Confirmation among reached: how clearly the product page stated price, delivery fee and delivery time; whether the ad promised something the page did not; the tone of the confirmation message.
• Completion: how long passes between confirmation and dispatch, whether the delivery address is complete, and who contacts the customer before the courier arrives.

Change one thing at a time and compare a comparable period. If you change the script and the offer in the same week, you will not know which one mattered.

A worked illustration of why it matters financially

Take the funnel above and suppose you spent 2,000 SAR on ads for those 100 orders. This is an illustrative example, not real results.

• Ad cost per placed order: 2,000 ÷ 100 = 20.
• Ad cost per delivered order: 2,000 ÷ 46 ≈ 43.5.

The platform reports the first number. Your bank account feels the second. If a change lifts delivered orders from 46 to 52 with the same spend, the cost per delivered order falls to 2,000 ÷ 52 ≈ 38.5, without touching a single ad. To turn this into a limit on what you can spend, start from the break-even calculation.

Trade-offs and when the advice changes

• Pushing harder can lower quality. Aggressive follow-up may raise confirmations while raising refusals at the door. Always read confirmation next to completion.
• Low volume makes rates noisy. With a few dozen orders a week, a swing of five orders is a big percentage. Look at a longer window before reacting.
• Mix matters. A product category, an ad source or a city may behave differently. Averages can hide a segment worth cutting or protecting.
• If your reach rate is already high, the best gain is probably after confirmation, in delivery, not in the call script.

Implementation steps

1. Write the four definitions above and agree them with whoever confirms orders.
2. Give each order a status that records how far it got: placed, reached, confirmed, dispatched, delivered, refused. Without statuses you cannot compute anything.
3. Record the time between each status; delay is often the hidden cause.
4. Report the four rates weekly, split by traffic source and product.
5. Pick the weakest step and test one change.

What you cannot know from this data

A confirmation rate does not tell you why a customer said no or never answered. It also cannot tell you which ad created the better customer unless the order record carries its source. Treat the numbers as pointers to where to ask questions, not as explanations.

Next step

If confirmations are handled by hand and statuses live in scattered chats, the measurement itself will be unreliable. COD Order Automation is built around recording and moving orders through these statuses. Then see what a failed order costs in Delivery and Return Rates: What a Failed Order Really Costs, and compare channel economics in COD vs Prepaid.