A Quote Is a Proposal Against Your Rules: Why Your Margin Already Leaked by the Time the Rep Names a Price

You closed the year with revenue up. Margin, not so much. You pull the report, and no one can point to exactly where it leaked. It wasn't one big, visible discount. It was thousands of small ones. Every rep, on every quote, deciding on the spot what was "okay" for that customer. Each call defensible on its own. Added up, they became the margin point that vanished from the books.

The easy explanation is "the market got tougher." Sometimes that's true. Usually it's half the story. The other half: your company never decided, centrally, what the pricing rules actually are. They exist. They just don't live in one place.

Why asking for a quote is really submitting a proposal

Treating a quote as "the customer asking for a price" hides what's happening. The customer is proposing a transaction: this product, this quantity, this company, this term, this region. Every variable triggers a rule. A price table by region. A credit limit per customer. Terms by channel and volume. Discount authority by management tier. A policy for slow-moving stock, a policy for the item that's out.

That proposal is evaluated against all of those rules at once. The problem isn't that the rules exist. It's where they live. One slice sits in the ERP, which was never built for the job. Another sits in the CRM. Most of it lives in the sales manager's head, learned by doing. Because they were never gathered in one place, they contradict each other: the credit rule fights the quarter's target, the regional table fights the deal a director cut with a big account.

Who reconciles that conflict, every time, on the spot? The rep. He becomes the human stitching scattered policies together to answer before the customer walks. It works. But it's expensive, and the cost is invisible, buried in one person's time and in the margin of a decision no one reviewed.

A quote doesn't ask for a price. It submits a proposal judged against rules nobody centralized.

Why margin leaks at the quote, not at the close

Because the margin decision happens before any report exists, the moment the rep says "I can do it for this." By the time the controller sees the number, the margin is already gone. The close only records it.

Follow the chain. Margin fell while revenue rose because the average discount crept up. It crept up because every rep decides on the spot, with no consistent ceiling by customer type. There's no ceiling because the rep doesn't have the rules at hand: credit, mix, item margin, the standing agreement. He doesn't have them at hand because they live in different systems and different heads. And they live that way because the company treated price as an individual talent of the sales team, not as company policy.

Root cause: the absence of a centralized pricing policy applied in real time. Not "a rep who discounts too much." That distinction changes who owns the problem. If the rep is to blame, the reflex is to push harder and tweak commissions: it treats the symptom and breeds conflict. If the problem is the architecture of the decision, the fix is to take the rule out of people's heads and put it where the quote happens.

The cost of not centralizing price

The bill doesn't sit in one line item. It spreads, which is why it goes unnoticed. This is the Cost of Inaction of the problem, on three fronts. Ungoverned discounting: each concession is small and defensible, the aggregate is the margin point that disappears. Reps as policy referees: your best seller spends part of the day reconciling rules instead of selling, back-office work disguised as commercial work. Margin decisions with no trail: no one can audit why a given discount went out, so no one learns, and the margin leaks again next quarter.

The economic name for this is transaction cost: the cost of making the deal happen, separate from the product itself. You can feel the order of magnitude. In a business doing $100M a year, one margin point lost to scattered discounting is $1M a year. Explicit assumption, illustrative number, not a measurement.

The reverse path has a measured size. Bain estimates that intelligent, data-driven pricing delivers roughly 3 points of margin, with a 5-to-11-point gap between companies that govern price and those that don't (Bain & Company, 2025). And it isn't a one-year fluke: McKinsey puts the sustained margin lift from pricing transformations at 2 to 7 points (McKinsey & Company, 2021).

From individual craft to governed system

This is where the pain meets the vision. For a long time, pricing in B2B was a craft. It lived in the experience of the person who'd been there twenty years and "knew" what each customer could bear. That knowledge is real. The problem is that it doesn't scale, can't be audited, and walks out the door when the person leaves.

The market's direction is clear: turn that craft into a system. Not to take the decision away from people, but to hand them the rule at the right moment. Pricing policy stops being folklore and becomes infrastructure: declared once, applied on every quote, auditable afterward.

This is where agentic commerce comes in, and it's worth framing carefully, because the topic is usually told wrong. The promise isn't an AI that sells on its own or pushes discounts. It's the opposite. With the rules centralized, an agent evaluates the customer's proposal against policy in real time and hands the rep a quote already calculated, inside the limits the company declared. The rep approves, adjusts, or rejects. Bain estimates that, applied well, AI can double the time a rep actually spends with customers and lift win rates by 30% or more (Bain & Company, 2025). And the size of the wave is already measurable: McKinsey projects agentic commerce could orchestrate $3 trillion to $5 trillion in global revenue by 2030 (McKinsey & Company, 2025). The machine carries the policy. The person owns the relationship.

Notice the prerequisite. None of this works on scattered rules. An agent bolted onto an ERP full of historical contradictions will execute the contradictions. Governance isn't an ornament you add after the AI. It's the condition for the AI to be trustworthy. First the policy becomes a system. Then the AI operates on it.

AI doesn't fix a pricing policy that doesn't exist. It only operates well on the rule a company had the discipline to declare.

How to get out of the pain

It doesn't start as a technology project. It starts as a management decision. Four principles.

1. Gather the rules in one place. Before any system, map where each pricing rule lives today and bring it into a single source. The exercise alone already exposes the contradictions.
2. Get the policy out of people's heads. Knowledge that exists only in the twenty-year veteran is risk, not advantage. Advantage is the rule that survives the person leaving.
3. Apply the rule where the quote happens. A policy in a PDF nobody opens governs nothing. It has to be at the point and the moment the price is set.
4. Leave a trail on every margin decision. Each discount with an author, a reason, and the rule applied. No trail, no learning.

Start small and free. Pull your last hundred quotes and answer one question: how many can you explain today by the rule that justified the price? If the answer is "depends on asking the rep," you've found where the margin leaks.

Frequently asked questions

Why does my margin fall even as revenue grows? Almost always because the average discount crept up in a scattered way: many orders with small concessions no one adds up. Revenue growth hides the erosion until it shows at the close, when the decision was already made earlier, at the quote.

Does centralizing pricing policy remove the rep's autonomy? No, it changes its nature. The rep stops guessing the ceiling and starts negotiating within clear limits, with the rule at hand. Autonomy moves to where it creates value, the customer relationship, instead of being spent reconciling policies that contradict each other.

Can more sales training fix this on its own? Training helps the symptom, not the cause. As long as the rule lives scattered across ERP, CRM, and people's memory, every rep rebuilds the policy their own way. The cause is architectural: the rule has to leave people's heads and move to where the quote happens.

On diagnosis, on prescription. The CWS Platform is built on exactly this idea: treating the quote as a proposal evaluated against declared, centralized rules applied in real time, with AI operating inside the limits the company set and the person in command. But the point of this piece isn't CWS Platform. It's that most of the margin leaking in B2B doesn't leak for lack of sales effort. It leaks for lack of a pricing policy that lives in one place. Recognizing that costs nothing. Continuing to pay the bill does.