COD order management: from order placed to cash collected

The states an order passes through

A COD order passes through about ten states, and each exists because a different decision is taken there. Collapsing them is the first cause of disorder: recording only "sent" or "not sent" no longer distinguishes a customer who needs calling back from a parcel waiting for collection.

What breaks when orders live in a spreadsheet

A spreadsheet holds the first few dozen orders perfectly well. It stops holding them not because of volume but because of the number of people: as soon as an agent, a packer and a courier touch the same order, the file becomes one copy among several.

• No single source of truth — The order exists in the spreadsheet row, on the courier's screen and in the memory of the agent who called. When the three disagree, nothing settles it.
• Re-entry — The same address is typed on the order, typed again for the courier, copied onto the label. Every copy adds its own typo, and a wrong address is a return you pay for.
• Statuses that contradict each other — The spreadsheet says "shipped" because somebody wrote it on Monday; the courier has said "returning" since Wednesday. The seller learns the difference when counting the money.
• Nothing records who did what — A spreadsheet overwrites the old value: you see the current address, never the one printed on the label.
• Profit known too late — A margin recalculated by hand arrives after the month is over, and nearly always overstated because one fee was forgotten.

What a single order record must hold

A useful record is not a complete one: it holds what is needed during a call, during packing and during a dispute, without opening anything else.

• The customer — Name, phone number and their other orders: someone already delivered to is not treated as a stranger.
• The address and the city — The city sets the delivery fee. Mistyped, it distorts the profit before the parcel leaves.
• The items — Exact reference and quantity: a size or colour variant is the leading cause of exchanges.
• The total the customer pays — The amount the courier must collect, the one quoted on the phone.
• The fees — Confirmation, the city's delivery fee, and the return fee where it applies — on the order, not in a separate calculation.
• The parcel — Courier reference and latest known status, pulled onto the order.
• The history — Every change of state with its date and author, plus the call notes.
• The profit — Derived from the lines above, never typed in by hand.

Why status history decides disputes

Disagreements in COD are about dated facts: the customer says they cancelled before dispatch, the courier says an attempt was made, the seller says nobody told them about the return. None of those sentences can be checked against a current state; they are checked against a sequence of timestamped ones.

That is why history must be appended and never overwritten. If the order keeps the time of the call, the amount agreed during it, the time the parcel was created and every courier status, the discussion ends in a minute; otherwise it ends in an estimate, and the seller pays for it. The same record makes failures readable: returns clustered on one city, one product or one lead source show up in the dates, not in a monthly total.

How profit is computed, and when it is credited

The profit on a COD order is only known on delivery, because collection is what creates it. Before that there is only an expected margin, and cash flow built on an expected margin eventually comes up short.

The calculation should be checkable line by line. On CODFamilia it is written with no percentage at all: seller profit = total paid by the customer − CODFamilia product price − confirmation fee − delivery fee. The confirmation fee is 10,00 MAD, charged once per order; the delivery fee depends on the city. No commission is taken on revenue. A seller who confirms their own leads pays no confirmation fee: they pay in telephone hours instead, which scales badly beyond a certain volume.

On delivery, the profit is credited to the seller's wallet; the seller then requests a withdrawal, subject to a minimum, and receives a supporting document. For orders already shipped but not yet delivered, an advance against profit can be requested, capped at a share of those profits and deducted from future ones: it is cash borrowed from yourself, and costly in clarity if it becomes permanent.

The reports that change a decision

A report is only worth having if it changes what you do tomorrow. Three questions are worth following, and all of them turn on the delivery rate: the share of dispatched orders the customer actually pays for.

1. Which products get delivered — A product ordered often and delivered rarely consumes advertising, delivery and returns while returning nothing. It should be withdrawn, not pushed harder.
2. Which cities cost money — Fees differ from one city to another and refusals are not spread evenly either. A city can be profitable on one product and loss-making on another.
3. Which lead sources hold up — Comparing sources on delivered orders rather than on leads received is what corrects an advertising budget.