Discount governance: who can approve what, and at what margin

Key Takeaways

• A company with no discount approval structure doesn't lose margin in one visible event. It loses it one small, defensible decision at a time, across every rep, every week.
• A tiered framework works on one principle: the size of the discount should set who has to see it, not who has to grant permission for it.
• The point is visibility, not control. Finance needs to see the pattern of discounting across the team to catch pricing problems early: a product priced wrong, a segment converting only at a discount, a competitor forcing concessions.
• Approval steps fail in two directions: too slow, and reps route around them; too loose, and the sign-off becomes a rubber stamp.
• The tiers only hold if the data behind them reaches somewhere finance actually looks, not individual deal approvals nobody aggregates.

A sales team without a discount approval structure doesn't bleed margin in one dramatic moment. It loses a point here to close a deal by Friday, another there because a competitor undercut the quote, another because a long-standing client asked and nobody wanted to say no. Each decision, alone, looks reasonable. Add them up across a quarter and a business can be giving away several points of gross margin with no single person responsible for the total and no record of where it went: research on B2B price waterfalls finds discounts, incentives and other off-invoice giveaways can account for up to a third of list price by the time a deal actually closes (McKinsey: The power of pricing, retrieved 2026-09-12).

A tiered approval framework fixes this not by taking discretion away from reps, but by attaching a visibility trigger to it. Below a certain size, a rep can grant a discount and move on. Above that size, someone with a wider view of the account or the margin structure has to see it before it goes out. The framework is less about permission and more about making sure the right person knows what's happening before it becomes a pattern instead of a one-off.

Why unstructured discounting bleeds margin invisibly

The reason this problem is hard to see coming is that no individual discount looks like a problem. A small concession to close a deal this month is a normal, defensible sales decision. The issue only becomes visible in aggregate: when someone pulls every closed-won deal from the quarter and finds the average selling price has drifted well below list, not because pricing changed, but because discretion did.

Without a structure, that aggregate view usually doesn't exist. Deals close in a CRM one at a time; nobody is looking at the discount field across the whole pipeline unless they go looking specifically for it, and research on discount discipline has found that fewer than a quarter of finance leaders actually track the total value of negotiated discounts given, let alone how they are distributed across segments and reps (Harvard Business Review: A Case for Discount Discipline, retrieved 2026-09-12). Finance sees a revenue number and a margin number, and by the time those two tell the story, the quarter that produced them is already closed. The rep who gave the discount isn't hiding anything. They were solving the problem in front of them, closing a deal, not protecting an aggregate margin figure nobody asked them to watch.

This is also where discounting becomes the path of least resistance for weak positioning. If a rep can't articulate why the price is what it is, the fastest way past an objection is to lower it. A team with no friction on that move will use it constantly, and the business will never find out that the product, the messaging, or the segment targeting was the actual problem, because the symptom got treated with margin instead.

A tiered approval framework: the logic, not a benchmark

The mechanics are simple, even if the exact thresholds a given business sets don't transfer cleanly from one company to another. The structure has three moving parts:

• A rep-level tier, where discretion is granted unilaterally because the discount is small enough that reviewing every instance would cost more in manager time than it protects in margin.
• A manager-level tier, triggered once a discount crosses a threshold where someone with visibility across the whole territory (not just this one deal) should weigh in before it's final.
• A senior or finance-level tier, reserved for discounts large enough, or frequent enough on one account, that they materially affect that period's margin and deserve a second set of eyes with a company-wide view.

Where exactly those thresholds sit depends on the business: margin structure, deal size distribution, and how much trust the sales team has already earned all move the lines. A services business with thin margins will set its tiers tighter than a high-margin software business closing annual contracts. There is no single percentage correct across industries, and a framework copied wholesale from another company's playbook will either strangle the sales team or do nothing at all.

Illustrative tier structure

The table below is one hypothetical shape, not a benchmark to adopt as-is. The right numbers come from a business's own margin data, not a template.

The profit margin calculator is a useful place to test where a given business's tiers should actually sit: running a handful of real deals through it at different discount levels shows how quickly a "reasonable" concession erodes the margin a deal was supposed to deliver.

This is about visibility, not distrust of the sales team

The framing that undermines most discount governance rollouts is treating it as a control problem: as if reps can't be trusted with pricing and need a manager standing over their shoulder. That framing is counterproductive: reps who feel policed tend to push back on the whole system or quietly find workarounds.

The actual purpose is aggregation. No individual rep can see whether the discounts they're granting fit a pattern across the team, because they only see their own pipeline, and discounting behaviour typically varies sharply between reps on the same team, with some giving concessions far more readily than others (Harvard Business Review: A Case for Discount Discipline, retrieved 2026-09-12). Finance can see that pattern, but only if the data reaches them in a usable form, not a hundred individual deal approvals sitting in a CRM, but a rolled-up view of who is discounting, how much, and on what kind of deal. A tiered structure generates that view as a byproduct of the approval workflow, rather than requiring someone to reconstruct it after the fact.

That distinction matters for how the framework gets introduced. A rollout pitched as "management doesn't trust you to price deals" invites resistance. One pitched as "finance can't see what you already know instinctively. That certain segments only convert with a discount, or one competitor keeps forcing concessions" tends to land differently, because it asks the sales team to help finance catch a problem they're already aware of, not to justify their judgment on every deal.

Where approval tiers fail: too slow, or too loose

Discount governance breaks in one of two predictable directions, and both are more common than a framework that simply works.

Too slow, and it gets bypassed. If a Tier 2 approval routinely takes three days because the manager who has to sign off is traveling or slow to respond, reps under pressure to close a deal will find a way around it: splitting a discount across two smaller approvals that each clear Tier 1, or absorbing the difference elsewhere in the deal terms. The moment an approval step costs more in lost deals than it protects in margin, it stops being followed, whether or not it's still formally policy.

Too loose, and it's meaningless. The opposite failure is an approval that exists on paper but functions as a formality: a manager who approves every request without reviewing deal context, because volume is too high or the incentive to push back is too low. A framework that generates approvals nobody reads produces the appearance of governance with none of the visibility it was built to create.

Both failure modes share a fix: the threshold has to match the approver's actual capacity, and the approver has to have a real stake in the margin outcome, not be added to the workflow as a formality. A framework that asks a manager to review forty discount requests a week fails the same way as one that asks them to review none. The thresholds also need revisiting on a real cadence (quarterly, against actual deal data) rather than left as policy nobody has looked at since it was written.

Discount governance sits inside the same discipline as the rest of a sales team's economics: what it costs to acquire a customer, what that customer is worth, and how pricing decisions made deal by deal add up to the number in the go-to-market and growth plan a business is actually running against. A tiered framework is a small piece of that discipline, but it decides whether the margin a growth plan assumes is the margin the business actually collects. Building and reviewing that tiered structure is easier from inside a sales accelerator that already tracks each rep's pipeline, so discount patterns surface next to deal velocity and conversion data instead of living in a spreadsheet nobody revisits.

Frequently asked questions

What size discount should trigger manager approval?

There is no universal percentage that applies across industries: a services business with thin margins will set the threshold tighter than a software business closing high-margin annual contracts. The right number comes from running real deals through your own margin data and finding where a concession starts materially affecting that deal's profitability, not from copying another company's policy.

Doesn't a discount approval process just slow down sales?

It will, if thresholds and turnaround expectations aren't matched to how the team actually operates. A Tier 1 limit set too low forces every small deal through review and invites bypassing; a Tier 2 or 3 approval with no committed turnaround gets routed around under deal pressure. Set correctly, the process only slows down discounts large enough that a delay is worth the visibility gained.

How is this different from just setting a maximum discount limit?

A flat maximum tells reps what they cannot do but gives finance no visibility into what they are doing below that ceiling. A tiered structure surfaces the pattern of discounting across the team (which segments, reps, or deal types are consistently discounting) so pricing problems get caught early, not just capped at an arbitrary limit.

The bottom line

Discount governance isn't a compliance exercise or a statement of distrust in the sales team. It's the mechanism that lets finance see a pattern reps are already living inside deal by deal but have no way to aggregate on their own. A tiered framework built around visibility, with thresholds set from real margin data and reviewed on a real cadence, is what keeps a company from discovering its margin problem only after the quarter that caused it has closed.

Figures were verified on 12 September 2026 against McKinsey and Harvard Business Review research on pricing discipline and discount governance. Specific tier thresholds are illustrative only and are not sourced, because the right numbers come from a business's own margin data, not a published benchmark.