Is Discounting a Sales Problem or a Pricing Problem?

Por Brian Tr · Atualizado em 10 de outubro de 2026

Every sales leader eventually gets the same email from finance: discounts are up, margin is down, and someone wants to know why reps keep giving money away. The reflex answer is a tighter approval policy. Add a VP sign-off above 15%. Require a written justification. Name and shame the worst offenders on the Monday call.

Our view is that this treats the symptom. Chronic discounting is usually a sign that the price doesn't match the value buyers can see, or that the package doesn't match what they want to buy. Reps discount because discounting works, and it works because nothing else in the deal moved the buyer. Approval rules can slow the leak. They don't find the pipe.

Key Facts

• In a 1992 Harvard Business Review analysis of 2,463 companies, Marn and Rosiello (McKinsey) found a 1% price improvement, with no loss of volume, raised operating profit by 11.1%, versus 3.3% for a 1% volume gain. A 1% price decrease would destroy the same 11.1%.
• A later McKinsey Quarterly article using typical S&P 1500 economics put the effect of a 1% price rise, with stable volumes, at an 8% increase in operating profit.
• In McKinsey's disguised lighting-supplier example, on-invoice discounts pushed invoice prices 32.8% below standard list prices, and off-invoice items took the average pocket price to about half of list.
• Individual transactions at that supplier sold at under 30% of list to 90% or more, a threefold spread for the same product.
• A study of a leading enterprise software vendor found reps agreed to significantly lower prices in quarters when closing a deal paid them more, costing the vendor roughly 6 to 8% of revenue (Larkin, Journal of Labor Economics, 2014).
• Simon-Kucher's 2025 Global Pricing Study, a survey of more than 2,200 business leaders across 28 countries and 39 industries, reports that companies realize less than half of their price increases on average.

Why the "discipline" diagnosis is attractive and usually wrong

The discipline story is comforting. It says the pricing is fine, the product is fine, and a few reps need to behave. It also gives you something to do by Friday.

But look at the arithmetic of the leak. The 1992 HBR analysis shows how lopsided price is as a profit lever: a 1% price move is worth about three times a 1% volume move. Every point a rep gives away to close a deal is a point you'd need roughly three points of extra volume to win back. Reps don't feel that exchange rate. They feel a buyer who says "I need 10% to get this through procurement" and a quota that closes Friday.

So the real question isn't why reps discount. It's why the discount is the only lever that moves the deal. If a rep could answer a price objection with a value case, a different package, or a different term, most would rather keep the price. Discounting is the path of least resistance, and a path that stays the path of least resistance is a design problem.

That's not an argument for removing controls. Controls matter, and a deal desk is the right place for exceptions. It's an argument for treating the controls as a measurement instrument first and a brake second.

Three root causes that look identical in the CRM

On the dashboard, all three show up as "average discount is too high." They need very different fixes.

1. Pricing and packaging. The list price or the package structure doesn't match how buyers value the product. Maybe the entry tier lacks the one feature every buyer asks for, so reps discount the next tier to compensate. Maybe the unit of pricing (seats, volume, modules) punishes your best-fit customers. The tell is that discounts cluster by package or segment, almost regardless of which rep sold it. Our piece on value-based pricing covers how to tie price to the outcome buyers care about.

2. Value selling. The price is defensible, but reps can't make the case. They lead with features, skip the economic buyer, or hand over a quote before the buyer has agreed what the problem costs. The tell is wide variation between reps selling the same package to the same segment, and discounts that appear late, after a stalled negotiation. A working value selling habit is the fix, along with coaching on handling price objections.

3. Process and incentives. Comp plans, quarter-end pressure, or a missing approval path make discounting rational for the individual. The tell is timing: discounts rise in the last weeks of a quarter, or in the quarter when a bonus accelerator is in reach. The academic record here is clear. Paul Oyer's 1998 paper in the Quarterly Journal of Economics found that, in many industries, average prices are substantially lower toward the end of fiscal years. Ian Larkin's enterprise software study found reps shifted deal timing and accepted deeper discounts in quarters where closing paid more.

Read the pattern before you pick the fix

Don't start with a policy. Start by cutting the last four quarters of closed-won deals four ways: by segment, by package, by rep, and by week of quarter. Where discounts cluster tells you what's broken.

Two warnings when you read this table against your own data. First, a pattern can have more than one cause, so treat the first fix as a hypothesis and re-cut the data after 60 to 90 days. Second, don't assume the extremes are errors. McKinsey's authors note that a wide pocket price band isn't necessarily bad, because customers and competitive situations differ. The question is whether the spread tracks something you can defend, like volume or strategic fit, or whether it tracks who negotiated hardest.

That second point is where most companies find money. In McKinsey's lighting-supplier example, many small buyers got far larger discounts than their volume warranted while some large buyers paid higher pocket prices. Nobody planned that. It accumulated one negotiation at a time.

What the leak looks like beyond the headline discount

Most leaders only track the discount field on the quote. The price waterfall is wider than that. The 1992 HBR article describes a pocket price waterfall, the path from list price down through invoice price to the price you actually keep, where each element (volume bonuses, payment terms, advertising allowances, freight) is a separate leak. In the McKinsey Quarterly lighting example, the 16.3 percentage points of reductions that didn't appear on invoices took the average pocket price to roughly half of list.

SaaS has its own version. Picture a 40-rep software company where the average quote discount is 12%. Add free implementation hours, an extended payment term, a price lock on renewal, and a "temporary" extra module. The real giveaway is closer to 20%, and none of it shows up in the field finance is watching. Illustrative, yes, but any revenue operations lead will recognize the pattern.

This matters for the diagnosis. If reps give non-price concessions freely, it often means price is hard to move, and the package doesn't flex in ways buyers value. That points toward packaging, not discipline.

What to do next

Resist the urge to publish a new approval matrix this week. A sequence that works better:

1. Instrument the leak. Track realized price against list at the deal level, including non-price concessions. If you can't see the full waterfall, you can't tell which of the three causes you have.
2. Run the four cuts. Segment, package, rep, week of quarter. Use the table above to form a hypothesis for each cluster. Bring finance and revenue operations to the same session, and the analysis will improve.
3. Fix the cause with the biggest dollar cluster first. If it's packaging, that's a product and pricing project with a quarter-long lead time. If it's value selling, it's a coaching program you can start Monday, with pipeline reviews that inspect deal quality rather than just stage and amount. If it's incentives, review the comp plan before the next planning cycle.
4. Use the deal desk for exceptions, not as a toll booth. Give it guardrails and a menu of trades (longer term, bigger commitment, narrower scope) so every discount buys something. The deal desk is also your best source of data on why discounts are requested.
5. Don't expect a price increase to stick without this work. Simon-Kucher's survey reports that companies realize less than half of their price increases on average. Discounting is one reason. Raising list into a leaky process mostly raises the starting point for the next negotiation.
6. Tie it to forecasting. Discount-heavy quarter-ends distort commit calls, which is one more reason to read this alongside forecasting discipline. And if reps sit in a structure that rewards volume over margin, sales org design is part of the answer too.

The test of whether you've fixed the cause rather than the symptom is simple. In six months, can reps hold price without needing an approver to tell them to? If yes, you've fixed the system. If margin only holds while the policy is being enforced, you've bought time.