Everybody Has the Price List. Almost Nobody Sells at It
· 7 min read · Written by Faceela Research & Editorial Team
Reviewed by Ahmed Hassan Algammal — Founder and Enterprise Systems Consultant
Ask a trading company what it sells an item for and you will be given a number, and the number will be honest. It is in the price list. It is on the website. Everyone in the building can quote it.
Now go and look at what came in.
The list price is the number the business names. The pocket price is what is actually kept once everything that came off it has come off — and in most trading businesses these are not two versions of one figure. They are two figures, held in different places, and only the first one has an owner.
Nothing about that gap is fraud, and nothing about it is even carelessness. Every layer that eats it was approved by somebody with the authority to approve it, for a reason that was good at the time. The problem is that no two of the layers are ever seen together.
The gap has named parts
Written down, the descent from list to pocket has seven steps, and a trading owner will recognise every one of them.
From the list to the pocket
Seven deductions, seven approvers, and no document that adds them
List price
The number in the price list. The only figure in this diagram that everyone in the business can name, and the only one nobody actually transacts at.
Customer discount on the price list
The standing percentage attached to the account, agreed once and inherited by every order after it. Nobody re-approves it, because approving it was somebody's job years ago.
Approved by: Whoever set the customer up
Line discount on the order
The one typed into the sales order for this deal, on this day, because the customer asked and the salesperson has authority up to a number.
Approved by: The salesperson, within his limit
Free goods
Eleven cartons invoiced, twelve delivered. It is a discount that leaves the price untouched, which is exactly why it is used, and it appears in the stock ledger rather than in the margin.
Approved by: The branch or sales manager
Extended payment terms
Ninety days instead of thirty. It is not on the invoice at all, and it costs the business whatever its own facility costs it for the extra sixty.
Approved by: Sales, and finance only afterwards
Freight absorbed
Delivered to the customer's site at our cost, agreed verbally, recorded as a distribution expense and never as a reduction of this customer's price.
Approved by: Logistics, or nobody
Rebate promised for the year
A percentage on volume, settled in twelve months. It is a real reduction of the price agreed today and it sits in a letter, not in the system.
Approved by: The owner or the commercial head
Credit note issued later
For a price dispute, a late delivery, a damage claim or a quality argument. It is priced against the original invoice, months after anyone was measured on that invoice's margin.
Approved by: Finance, resolving somebody else's promise
Pocket price
What was actually kept, against the landed cost of the item rather than the purchase order price. This is the only figure on the diagram that is the margin — and it is the one figure no report in the business produces, because producing it means one screen showing all seven layers against one customer, which is the screen the seven approvers have never had to look at together.
Read the approver column rather than the layers. Seven deductions from one price, seven different people, and not one of those seven ever sees what the other six did. The customer discount was set by whoever opened the account. The line discount is the salesperson's. Free goods are the branch manager's. Terms are negotiated by sales and paid for by finance. Freight is absorbed by logistics or by nobody at all. The rebate is the owner's, in a letter. The credit note is finance, months later, settling a promise somebody else made.
That is not a control failure. There is no control there to fail.
Approval against nothing is not approval
The mechanism is worth being precise about, because it explains why tightening any single layer never works.
Each approver is asked a question that is locally reasonable. Can we give this customer another two points to close it? The answer is judged against the item's standing margin, which the approver reads off a screen, and which is calculated from the purchase order price rather than the landed cost.
So two separate errors compound. The margin being protected is overstated, because freight and duty are not in it. And the deduction being approved is one of seven, six of which are not on the screen. A discount waved through as still twelve points can be three by the time landed cost is applied and the other layers are counted — and nobody involved did anything wrong, because nobody was ever shown the total.
The correction is not stricter approval. It is that the approval screen has to show the pocket price for that customer, on that item, including everything already given, against landed cost. One number, at the moment of the decision. Everything else is a policy nobody can execute.
Authority is a matrix, not a percentage
Most discount policies are a single number per role — the salesperson has five points, the manager has ten. It is easy to write, easy to configure, and it controls the one layer that is easiest to see while leaving the other six untouched.
The salesperson who cannot give another point can still give ninety-day terms, or a free carton, or delivery to site. None of those touch the discount field. All of them reduce the pocket price. A policy expressed as a percentage of one layer is an invitation to move the concession into a different layer, and that is exactly what happens, without anyone intending it.
Authority has to be expressed per layer per role instead:
| Concession | Salesperson | Branch manager | Commercial head | Owner |
|---|---|---|---|---|
| Line discount | Within a standing limit | Above it, to a second limit | Above that | Anything, logged |
| Free goods | No | Yes, against a monthly budget | Yes | Yes |
| Extended terms | No | Referred to finance | With finance | With finance |
| Freight absorbed | No | Yes, priced onto the order | Yes | Yes |
| Rebate commitment | No | No | Yes, in writing | Yes |
| Credit note for a price dispute | No | Raise only | Approve | Approve |
The grid matters more than the values in it. Whatever your own limits are, the point is that every column of the pocket price waterfall has a row here, so a concession cannot be moved sideways into a layer that nobody governs.
Notice the last row in particular. A credit note for a price dispute is a retrospective discount, and in most trading businesses it is the only layer with no approval rule at all, because it arrives dressed as a customer service matter.
Four prices, one customer, three branches
The symptom that finally makes an owner ask the question is usually this one.
The same customer, buying the same item, in the same month, from three branches, at four different prices. Nobody lied to him. Each branch quoted honestly from what it could see, applied the discount its own manager had authority to apply, and closed the sale.
The reason it happens is structural rather than behavioural: pricing is held per branch, or per salesperson's spreadsheet, or per the last invoice anyone can find, and there is no single answer to the question what does this customer pay for this item? That question has to have exactly one answer, produced by the system, at the moment of quotation, or it will have as many answers as there are people who can raise a quotation.
And when the customer notices — and he will, because his three purchasing officers talk to each other — you lose the highest of the four prices, permanently, on every branch.
What to make a vendor show you
On a live system, with a real customer and a real item, not on slides.
- Show the pocket price for one customer on one item, with all seven layers itemised, on one screen.
- Show that number against landed cost, not against the purchase order price.
- Refuse a line discount above a role's authority — a refusal, not a warning — and route it for approval.
- Give free goods and show the pocket price move, in the same place the discount would have moved it.
- Record extended terms and show the financing cost land on the order, not in a finance report nobody links back.
- Produce realised price for one customer across every branch and every channel for a past quarter, ranked by variance.
- Show a credit note issued for a price dispute reduce that customer's realised price for the original period.
Item 1 decides it. Every system can calculate a discount. A system that can show all seven layers against one customer on one screen has done the only thing that makes the other six items possible — and it is the screen that no approver in most trading businesses has ever been shown, which is why the approvals have been reasonable and the outcome has not.
The short version
The list price is the number the business can name. The pocket price is the number the business keeps, and they are separated by seven layers approved by seven people against no total.
Governing the discount percentage alone moves the concession into free goods, into terms, into freight and into a credit note next quarter. Authority has to be a grid covering every layer, or the layers that are not covered become the ones that are used.
And the margin every one of those decisions is judged against is the wrong margin to begin with, because it is calculated on the purchase order price rather than on what the goods actually cost.
Which makes this article and landed cost the same problem seen from the two ends: the cost is understated on the way in, and the price is overstated on the way out, and the gross margin report is arithmetically correct about neither. The suite that puts the pricing matrix, the approval routing and the landed cost on one record is Logix. Turning the approval into something that survives a busy quarter is process automation, and the wider case is ERP for trading and distribution in the UAE.