A policy is what keeps one-off negotiation from becoming the price
Knowing how to negotiate a single deal is a rep skill. Making sure thirty deals a quarter don't each quietly leak ten percent is a team problem, and skill doesn't solve it — policy does. The difference matters because the damage from undisciplined discounting isn't one bad deal; it's the slow erosion that happens when every rep, under pressure to close, reaches for the same easy lever, and nobody can see the pattern until the average selling price has drifted down a quarter at a time.
Without a policy, the discount becomes the default move. A rep hits a stall, knocks off fifteen percent, closes, and gets rewarded — so they do it again, and the rep next to them learns the same trick, and within two quarters your "price" is whatever the buyer pushes it to. A discount policy exists to make holding price the easy path and giving it away the deliberate, visible exception. It's not bureaucracy; it's the thing that lets reps move fast on the deals that deserve a concession without turning your list price into fiction.
Decide what a discount is for before you set the rules
A policy that just caps the percentage misses the point. The first question isn't "how much can a rep give away" — it's "what is a discount supposed to buy?" A discount is a tool to win something, and a policy that treats it as a tool produces very different behavior than one that treats it as a ceiling.
The healthy answer, the same logic behind trading concessions instead of giving them away, is that a discount should always purchase something for the business: a longer term, more volume, faster signature, a reference, a multi-year commitment. A policy built on that principle doesn't say "you may discount up to X." It says "you may discount in exchange for these things," which reframes every concession as a trade and quietly kills the reflexive giveaway that has nothing on the other side of it.
Build approval tiers that match the cost of the decision
The core mechanism of any workable discount policy is a tiered approval matrix, and the art is setting the thresholds so the policy protects margin without becoming a bottleneck that pushes reps to route around it.
A simple, durable structure looks like this:
- Rep discretion (small). Up to a modest threshold, the rep decides alone. This has to be generous enough that the everyday "round it to a clean number" concession doesn't require a meeting — otherwise reps learn to resent and dodge the whole system.
- Manager approval (meaningful). Past that line, a manager signs off. This is where most real negotiation lands, and the approval step is less about saying no than about forcing the rep to articulate what the discount is buying.
- Leadership approval (strategic). The deep cuts — the ones that set precedent or torch margin — go up a level, because their cost isn't just this deal; it's every renewal and reference that follows.
The thresholds matter less than the principle: the size of the discount should match the seniority of the person approving it, and crossing a tier should require a reason, not just a signature. A policy that asks "what did we get for this?" at every tier turns approval from a rubber stamp into the moment the trade gets made real.
Floors, not just ceilings
Most policies set a maximum and stop there. The stronger ones also set a floor — a margin or price below which the answer is simply no, regardless of who's asking or how badly the rep wants the logo. The floor is what protects you from the deal that closes at a number you'll regret at renewal, and from the rep who's convinced this marquee account is worth selling at a loss.
The floor also does something subtler: it gives reps cover. "I genuinely can't go below this — it's a hard floor, not me being difficult" is a far stronger position than a rep improvising where to stop. A real walk-away point, decided in advance and backed by policy, is exactly the negotiating leverage a rep needs, and a healthy pipeline is what makes honoring the floor possible — you can decline the bad-margin deal because you have others.
A policy nobody can see isn't a policy
The most common way a discount policy fails isn't bad thresholds — it's invisibility. If discounts live in reps' heads and PDFs and one-off Slack approvals, you can't see the pattern until renewal, when you discover half the book is priced ten points under list and nobody remembers why. A policy you can't measure is a policy you can't enforce.
This is where the discipline of logging the real economics of every deal stops being hygiene and becomes the thing that makes the policy real. If every concession is recorded as a trade — the discount given and the term, volume, or commitment received for it — then discounting stops being invisible and starts being a pattern you can manage.
Make the policy live where the deals do
A discount policy enforced by memory and goodwill decays the moment the quarter gets tight. The policy has to live in the system the deals already move through, or it's just a document.
In Hitt CRM, the deal record holds the actual value, the discount applied, and the concession traded for it, so every cut is documented as a trade rather than forgotten as a giveaway — and an automation can route a deal that crosses a discount threshold to the right approver as a task, so the tier matrix runs itself instead of depending on a rep to remember to ask. Reports then turn a quarter of deals into the answer the policy exists to produce: who's holding price, where margin is leaking, which segments push hardest, and whether your average selling price is drifting. The policy stops being a PDF nobody reads and becomes a guardrail the pipeline enforces in the flow of the work.
The one-sentence version
A discount policy protects margin not by capping the percentage but by deciding what a discount is for — a trade for term, volume, or commitment — then enforcing it with approval tiers sized to the cost of the decision, a hard floor that gives reps cover to walk, and full visibility in the CRM, so that holding price is the easy default and giving it away is a deliberate, recorded, paid-for exception rather than the quiet way your list price erodes a quarter at a time.