COD Order Profitability

Cash on delivery is usually discussed as a conversion decision — offer it and more people buy. It is also a margin decision, and the two point in opposite directions.

The reason it gets misjudged is that a delivered COD order genuinely looks good. There is no card processing percentage, which on a $92 order saves nearly three dollars against a card payment. Judged on delivered orders alone, COD can appear to be the better payment method.

The orders that never deliver are where the economics actually live.

What a failed attempt costs

An undelivered COD parcel produces no revenue and consumes:

• Outbound freight, already paid.
• Return-to-origin freight, billed separately and usually at a similar rate.
• Packaging, consumed and rarely reusable.
• Handling and restocking at the warehouse.
• A failed-attempt or COD handling fee from the carrier, depending on contract.
• Inventory tied up in transit for the full round trip, unavailable to sell.

Cost of goods usually returns to stock, which is the one saving. Everything else is spent.

In the worked example, a failed attempt costs $19.35 — against a delivered order that contributes $46.25. Roughly two failures wipe out one success.

Expected value is the number that matters

The mistake is treating failures as an occasional annoyance rather than as a rate that applies to every order you accept.

You pay acquisition cost on attempted orders, not delivered ones. So the figure to compare against CAC is expected contribution per attempt:

(delivery rate × margin when delivered) − (failure rate × cost of a failure)

At an eighteen percent failure rate, that turns $46.25 into $34.44 — a quarter lower. A store bidding against the $46.25 figure is overbidding by nearly twelve dollars per customer, on a method chosen partly because it seemed cheaper than cards.

Recomputing this is the single highest-return hour available to any store with meaningful COD volume.

Where failure concentrates

Failure rates are never evenly spread, and the variation is where the fixes are.

By region. Certain postcodes fail far more than others, for reasons ranging from address quality to carrier coverage to local delivery practice. This is usually the largest cut.

By order value. High-value COD orders fail more often, because the customer has to produce more cash at the door and second thoughts are more expensive.

By product category. Impulse categories fail more than considered purchases.

By acquisition channel. Traffic from aggressive discounting converts more casually and refuses at the door more often.

A store-wide failure rate averages all of this into one uninformative number. Cut by region and by order value first — between them they usually explain most of the variance.

The levers, and what each costs

Every intervention trades conversion for reliability, and the trade is measurable.

Order confirmation before dispatch, by message or call. Reduces failures materially, adds a step, and loses some orders that would have delivered.

Partial prepayment — collecting a small deposit online. The most effective single lever, because it converts a costless refusal into one the customer has already paid something to avoid, and it loses the customers who wanted COD precisely because they were not committing.

Restricting COD above a value threshold or in specific postcodes. Precise, easy to implement, and the finding from the regional cut usually points straight at the list.

Address verification at checkout. Cheap, and it removes the failures caused by bad data rather than by intent.

The right combination is a calculation, not a preference: compare the margin lost to reduced conversion against the failure cost avoided, using your own numbers.

Reading COD alongside prepaid

Because the two payment methods have genuinely different cost structures, blending them produces a figure that describes neither.

Split contribution margin by payment method and read four numbers together: margin when delivered, failure rate, expected margin per attempt, and share of orders. That combination answers the question stores actually have, which is not “is COD good” but “how much COD can we carry”.

The pattern that usually emerges is that COD is fine at the volumes and values where it started and deteriorates as the store grows into higher-priced products. A method that worked comfortably on $40 orders can be marginal on $150 ones, because the cost of a failure scales with freight and handling while the customer’s cost of refusing stays at zero.

That framing also makes partial restrictions easy to justify. Keeping COD on lower-value orders while requiring prepayment above a threshold is not a rejection of the method — it is applying it where the expected value works and withdrawing it where it does not. The threshold comes straight out of the numbers rather than from a policy debate.

Working capital, which nobody counts

One cost sits outside the margin calculation entirely and belongs in the decision.

With COD, money arrives when the parcel is delivered and the carrier remits — commonly a week or two after dispatch, sometimes longer. With prepaid orders, it arrives in a payout a few days after checkout.

For a store funding inventory from its own cash flow, that gap is a real constraint. Growing COD volume means more money in transit at any moment, and the faster you grow the wider the gap. Stores that hit a cash wall while reporting healthy margins are frequently doing so for exactly this reason: the profit is real, and it is currently in a van.