From order to parcel: the actions at the packing table

The article beside this one argues that an order is not a sale. It covers what happens before — the channel, prepayment, geography — and what happens after: the refusal, the return, the parcel that comes back.

Between the two there is a table. On that table an order becomes a physical object carrying an address, and how that happens decides part of what the neighbouring article describes as a delivery problem.

That zone is almost never written down. It is made of actions learned on the job, different depending on who is there, and repeated slightly differently every day — which is exactly the definition of a place where errors are normal.

We describe the actions in order: the table, the three gates an item passes, what is written on the box, counting what moves, and the cost of an error — which exists, which is recorded, and which is recorded somewhere else.

An item passes three gates between the shelf and the sealed parcel, and each is a moment when it can still be recovered.

The first is picking: fetching the right item, in the right quantity, in the right variant. That is where reference errors happen, and they almost always come from two neighbouring items that resemble each other.

The second is checking: verifying that what is on the table matches what is written. It is the gate most often deleted, because it looks redundant when you have just picked the items yourself — and the next section explains why that impression is false.

The third is closing: the parcel is sealed, the address applied, the order marked as packed. The order of those three matters: marking before sealing produces orders marked ready that are not.

A gate passed is not reopened silently. If a sealed parcel has to be opened — a forgotten item, a doubt — it goes back through the check before being resealed, or the verification that took place is worth nothing.

The address on a parcel is the leading cause of delivery failure in this country, and the neighbouring article gives it a whole section. The table is where it is written, and therefore where it is corrected.

Three elements beat a long address: the name, the phone number, and a landmark. The landmark is what the courier actually uses — a shopfront, a junction, a mosque, a building — and it is worth more than a street name nobody uses.

The phone number has to be legible without effort and checked at packing. An unreadable digit produces a call that does not connect, and a call that does not connect produces a parcel that goes back — the highest cost in this whole chain.

What is not written matters too: the nature of the contents, their value, and your brand name if the product is of a kind that interests people. A parcel announces its contents to everybody who handles it, and you know none of them.

Finally, the label goes on flat and on one face, never across an edge. A folded label becomes illegible after three handlings, and the address underneath it no longer exists.

When payment happens on delivery, the table carries an extra responsibility: the amount to collect travels with the parcel, and an error in it is discovered in front of the customer, by somebody who is not you.

So the amount is checked at closing, like the reference and the quantity, and it is checked against the order rather than from memory. It is a fourth line on the check, and the only one that translates directly into money.

The second point is consistency between what the customer understood and what is written. A discount granted in conversation and absent from the amount to collect produces an argument on the doorstep that the courier cannot arbitrate.

The third is change. An amount that forces the courier to give change on a large note is a real source of failure in this payment method, and rounding is not always possible — but knowing it at least allows the customer to be warned.

Finally, the table has to know what was collected and when. Remittance is handled in the neighbouring logistics article; what concerns the table is simpler: a column, a date, an amount, and reconciliation done weekly rather than quarterly.

What we do is short and physical: we write the page of actions with you, we put the picking slip into the right format — no price, tickable line by line — and we establish the short list of references to count each week.

We do not sell warehouse software to a business packing thirty parcels a day. A square metre, three strips of tape, a printed slip and one rule about open orders remove most errors, and no tool will do that work instead of the hands.

We do not pack your parcels, and we will not recommend outsourcing this step while the actions are unwritten. Outsourcing a process that does not exist means buying somebody else’s errors on top of your own.

We will promise no quantified reduction in the error rate, for the reason written in the previous section: the starting cost is spread over four accounting lines that do not carry that name, and an improvement figure built on that would be a fabrication.

Finally, all of the above happens without us. One open order at a time, a three-point check, a landmark rather than a street, and a short list counted on Monday: four free decisions, and they are worth more than the rest of this article.