The Five Numbers That Actually Run a COD Business

Why revenue is the wrong headline number in COD

In prepaid e-commerce, revenue is money you already hold. In cash on delivery it is a forecast, and often an optimistic one. An order that is confirmed but never delivered still shows up in most dashboards as a sale, right up to the moment the parcel comes back and quietly reverses it three weeks later.

That lag is what makes COD businesses misread themselves. A month that looked strong on order count can turn out to have been the most expensive month of the year once return freight lands. The way out is to stop watching the top line and start watching the handful of ratios that predict what actually gets collected.

The five numbers worth reviewing every week

Every one of these is a ratio, not a total, which is what makes them comparable across a slow week and a peak campaign. Track them on a fixed cadence and the trend tells you more than any single value.

• Confirmation rate: confirmed orders divided by inbound conversations or leads. It measures how well you turn interest into a real, dispatchable order.
• Delivery rate: delivered orders divided by shipped orders. It measures how well your address data, expectations and reminders survive the last mile.
• Return to origin (RTO): parcels that come back divided by parcels shipped. It is the inverse of delivery rate and it is where the money burns twice.
• Cost per delivered order: total spend, meaning ads plus shipping plus return freight plus confirmation labour, divided by delivered orders. Not per order taken. Per order actually paid for.
• Contribution margin per delivered order: what is left after product cost and every variable cost above. This is the only number that tells you whether scaling helps or hurts.

The numbers only make sense together

Read in isolation, each of these can lie. A confirmation rate that jumps after you loosen your qualification is not good news if delivery rate falls by more. A delivery rate that looks excellent may simply mean you are refusing to ship to the regions where growth is.

The honest view is the funnel end to end: how many conversations became confirmed orders, how many of those shipped, and how many turned into collected cash. Every stage compounds, so a modest gain early is worth more than a heroic effort late.

What each number tells you to fix

A weak confirmation rate is almost always a response-time problem before it is a persuasion problem. Look at how long it takes to answer a first message at 10 PM, not at your best-case daytime average. Then look at how many quiet conversations ever get a second touch.

A weak delivery rate is an information problem. Wrong phone digits, incomplete addresses, buyers who were never told when the parcel was coming. These are captured or lost in the conversation, days before a courier is involved, which is why fixing them is a messaging job rather than a logistics job.

A cost per delivered order that keeps climbing while confirmation and delivery hold steady means the leak is upstream, in traffic quality. You are paying for clicks from people who were never going to complete an order, and no amount of downstream discipline fixes that.

Where the measurement usually breaks

Most COD sellers cannot compute these numbers cleanly, not because the maths is hard but because the data lives in three places: an ad account, a chat inbox and a carrier portal. Orders are re-typed between them, statuses are updated by hand and late, and returns are reconciled once a month if at all.

The fix is structural. When conversations, orders and carrier tracking share one pipeline, every stage of the funnel is timestamped automatically and the ratios become a report rather than a research project. That is the practical reason to keep the sales conversation and the order record in the same system: not tidiness, but measurability.

Frequently Asked Questions

What is a good COD confirmation rate? It varies widely by product and traffic source, so the useful benchmark is your own trend. A steady improvement on your baseline is worth more than any industry average.

Is RTO the same as a return? Not quite. RTO is a parcel that never reached the customer and came back, while a return happens after a successful delivery. RTO is more expensive because it earns nothing at all.

How often should I review these? Weekly is enough for the ratios and monthly for contribution margin, which needs returns to settle before it is accurate.