Customer price lists, volume breaks, promotions and negotiated prices: where each one comes from, the order they apply in, and what happens when the credit limit is used up.
by Nadia Maar ·
The customer calls and says last time they paid 4.20. Inside the company three people look in three different places and give three different numbers: the price list says 5.10, the contract says 4.60, and the last invoice really does say 4.20 because there was a promotion in March.
All three are right. The problem is not the commercial policy — that is rich because the market demands it — it is that nowhere is it written which rule wins when two apply at once. The order of precedence exists, but it lives in the head of whoever has been there longest, and it changes with their mood.
This piece is that order, written down. Plus the second half of the same problem, which almost nobody connects to the first: the credit limit. Because a price agreed with someone who cannot pay is not a sale, it is a loan.
In a B2B relationship the price of a product is not a number: it is the result of six sources living side by side, each created for a sensible reason and at a different moment.
None of these is wrong. The mistake is thinking they can live in a single table: they are rules that all apply at once, and the work is not picking one, it is deciding what order they apply in.
This rule comes before all the others, and it is technical rather than commercial. If the site calculates the price with its own logic and the back office recalculates with its own, sooner or later the two results diverge — and the discovery happens on the invoice, which is the worst possible moment. Only two architectures hold: one calculation engine used by both, or the back office deciding and the site asking. Everything else is a manual reconciliation dressed as an integration.
The six rules are always there; what changes is how many apply. Below, the same product goes through three different situations: the rules that lose stay visible with their reason, and that is the part that matters. Switch customer.
VT-4180 · 18V battery driver · order of 12 units unless stated
| Base price list 2026 | 5.10 € | Nothing else applies: it is the public price, and for a customer with no history it is the right one. |
| Customer tier | — | The customer has not been assigned a tier yet. Tiers are given once there is a history, not at registration. |
| Negotiated price | — | No contract in place on this product. |
| Volume break | — | The order is 12 units; the first break starts at 50. |
| Active promotion | — | No promotion on this product at this date. |
| Rep override | — | No rep assigned to this customer. |
Price applied: 5.10 €. The simple case, and it is here to show the structure: all six rules exist, five do not apply. A system that shows just «5.10» here is perfectly fine; a system that cannot say why starts being a problem at the second case.
| Base price list 2026 | 5.10 € | It remains the basis for percentage discounts, but it is not the price. |
| Customer tier (−8%) | 4.69 € | It beats the list price but loses to a more specific rule. Specificity is the criterion: the rule that names this customer and this product wins. |
| Negotiated price | 4.60 € | A number agreed for this product with this customer. It is the most specific rule there is, which is why it sits at the top of the order. |
| Volume break | — | The order is 12 units. |
| Active promotion | — | The March promotion expired on the 31st. And it genuinely ended, because it had an end date written on it. |
| Rep override | — | Not requested. |
Price applied: 4.60 €. And above all: you can tell the customer WHY it is 4.60, in one line, without calling anyone. That is the difference between a price list and a mystery.
| Base price list 2026 | 5.10 € | Basis for calculation, not a price. |
| Customer tier (−8%) | 4.69 € | Beaten by the more specific rules. |
| Negotiated price | 4.60 € | This is where the most argued-about question in any price list has to be settled once: does the volume break beat the negotiated price, or not? In this price list it does, because the contract sets a ceiling and not a floor. The opposite answer is equally valid — what is not valid is not having written it down. |
| Volume break (100+) | 4.35 € | The order is 120 units. The quantity discount is calculated on the negotiated price, not on the list price: stacking discount on discount from the list price is the fastest way to give away margin without noticing. |
| Active promotion | — | No promotion running. |
| Rep override | 4.20 € requested | It exceeds the ceiling allowed for the role, which is 5% off the resulting price. It is not applied automatically: it waits for approval, and the person who asked can see it is waiting instead of hearing it from the customer. |
Price applied: 4.35 €. The two losing rows here are exactly the ones that generate the phone calls: without a written order, both 4.60 and 4.20 are defensible, and the argument is won by whoever is louder.
The document you need is not technical and it is half a page long: the six rules in order, and for every pair that can collide, which one wins. It gets written in one meeting, argued over for an hour, and removes a year of later arguments.
The second half of the problem is that the right price is worth little if the customer can no longer buy. And «how much is already committed» is a sum of four things, not just the overdue amount.
| Invoiced, not yet paid | €14,200 |
| Delivered, not yet invoiced | €4,800 |
| Orders accepted, not shipped | €3,200 |
| Overdue beyond 30 days | €1,200 |
| Incoming order | €2,900 |
limit · €25,000
The incoming order takes exposure to €26,300, which is €1,300 over the limit. All three possible answers are legitimate — block it, accept it with someone's approval, or simply accept it — and they change from customer to customer. What is not legitimate is finding out at month end: the two middle bands, delivered-not-invoiced and accepted-not-shipped, are the ones almost no system adds up, and here they are worth eight thousand euros.
In all three cases the thing that changes the outcome is the same: exposure is calculated at the moment the order is being composed, not the following night. A block arriving twenty-four hours later has prevented nothing — it has only added an awkward phone call to an order already promised.
The temptation is to start with the back-office integration, because it is the part that looks technical and therefore urgent. It is the wrong order: until the six rules have a written precedence, the integration simply moves the ambiguity between systems faster than before. The half-page document first, then the code — and the document is written by the commercial people, not by whoever develops.
In the audit we take one of your customers and one of your products and rebuild the price together: which of the six sources apply, in what order, and whether the number that comes out is the one that ends up on the invoice. In most cases at least one pair of rules turns up whose precedence nobody ever decided.
Then we look at the credit limit the same way: what your exposure figure adds up today, and what it does not. If it turns out you already have all of this in order, I say so and we finish before the half hour is up.
Not with a price table, but with an order of precedence between rules that all apply at once: base list, customer tier, negotiated price, volume break, time-limited promotion and rep override. The most specific rule wins — the one naming this customer and this product — and every pair that can collide needs a precedence decided in writing. Without that document, the conflict is settled by whoever is loudest.
There is no universal order, only the order you chose and wrote down. The most argued-about question is always the same: does the volume break beat the negotiated price, or not? Both answers are defensible — it depends whether the contract sets a ceiling or a floor — and what is not defensible is not having decided. One technical rule is clear-cut, though: the quantity discount is calculated on the resulting price, not on the list price, because stacking discounts gives away margin unnoticed.
By adding four things, not just the overdue amount: invoiced but unpaid, delivered but not yet invoiced, orders accepted but not yet shipped, and the genuinely overdue balance. The two middle items are the ones almost no system adds up, and on an average customer they are often worth several thousand euros. The calculation has to happen while the order is being composed: a block that fires the following night has prevented nothing.
It never should, and it is a question of architecture before configuration. If the site calculates with its own logic and the back office recalculates with its own, sooner or later they diverge, and the discovery arrives on the invoice. Only two solutions hold: one calculation engine used by both, or the back office deciding and the site asking. It also helps to freeze the price when the order is accepted, so a price list updated the following week does not change an order already taken.