---
title: "Pricing audit: where discounts eat your margin | MarginFix"
description: "A pricing audit checks whether each price, discount and promotion still covers its full cost. Mispriced offers turned up in 47% of our 74 audits."
url: https://marginfix.ai/insights/pricing-audit/
lastmod: 2026-10-12
---

[Home](/) › [Insights](/insights/) › Pricing audit: where discounts eat your margin

Pricing · 8 min read

# Pricing audit: where discounts eat your margin

![Portrait photograph of Cristian Bragau](/assets/founder/cristian-bragau-2026-10.jpg) **[Cristian Bragau](/about/cristian-bragau/)**, Founder & Lead Auditor · Reviewed Oct 2026 · verified against the base of 74 audits

The short answer

A pricing audit tests whether every price, discount, rebate and promotion still covers its full cost and earns the margin it was set to earn. In our 74 audits, mispriced offers appeared in 47% and discount habits in 34%. Most of the leak hides in discounts that were set once and never revisited.

47% of audits found mispriced offers

34% of audits found a discount habit

$9 of every $100 of audited spend went to mispriced offers

## What a pricing audit checks

A price list is the visible half of pricing. The other half is everything given away between the list price and the margin you bank: promo codes, free shipping thresholds, bundles, plan discounts, dealer and channel discounts, rebates, fee waivers, and surcharges that never made it into a contract. A pricing audit tests each of those against its full cost today, not against the cost of the year it was set.

Each one usually started as a sound decision. A launch discount won the first customers, a threshold matched the carrier rates of its year, a rebate tier fit the distributor it was written for. Then freight, materials, platform fees or the customer mix moved, and nobody went back. That is why mispriced offers appeared in 47% of the 74 audits and discount habits in 34%, and why the leak rarely shows on the price list itself.

## Where discounts eat margin

Across the published cases, the pricing leaks fall into five places. Each comes with the test that finds it.

### Codes that became the price

A code offered every week stops being a promotion and becomes the real price, and the customers most likely to pay full price learn to wait for it. [A discount habit was training the best customers to wait](/cases/a-discount-habit-was-training-the-best-customers-to-wait/) is that pattern in a food and beverage brand. Bundles can do the same from the other side, pulling buyers off higher margin single items: [Bundles were cannibalizing full price hero SKUs](/cases/bundles-were-cannibalising-full-price-hero-skus/). The test is the share of orders that carried a code, month by month, against full price demand over the same months.

### Thresholds and extras priced years ago

Free shipping thresholds, free delivery and free assembly are prices too, set against a basket and a carrier rate that may no longer exist. [Free shipping was quietly deleting the margin on every third order](/cases/free-shipping-was-quietly-deleting-the-margin-on-every-third/) is the threshold case, and [Free assembly and delivery were bundled below cost on big ticket lines](/cases/free-assembly-and-delivery-were-bundled-below-cost-on-big-ti/) shows the same leak on big ticket furniture. The test is the contribution of every order just above the threshold, after the shipping it now costs.

### Plan, deal and channel discounts

Discounts written into plans, deals and channels outlive the reasons they were granted. An annual plan discount is worth its cost only if it prevents more churn than it gives away, and in [Annual plans were discounted below their retention value](/cases/annual-plans-were-discounted-below-their-retention-value/) it did not. A dealer discount applied across a whole catalog subsidized the low margin lines in [Distributor margin was set once and never revisited](/cases/distributor-margin-was-set-once-and-never-revisited/), and discounts granted to close deals before quarter end compounded in [Late quarter discounting was quietly eroding ACV margin](/cases/late-quarter-discounting-was-quietly-eroding-acv-margin/). The test is a list of every standing discount with its date and its reason, and a check that the reason still holds.

### Rebates and waivers nobody switched off

Rebates and fee waivers are discounts paid after the sale, so they rarely show up in a pricing review. Distributor rebates were accrued on orders that never reached their tier in [Volume rebates were paid on orders that never hit the tier](/cases/volume-rebates-were-paid-on-orders-that-never-hit-the-tier/), and launch era fee waivers kept suppressing a marketplace’s [take rate](/glossary/take-rate/) long after launch in [Promotional fee waivers were never switched off](/cases/promotional-fee-waivers-were-never-switched-off/). The test is one quarter of rebates and waivers recalculated from shipped and paid volume, then compared with what was credited.

### Prices below the full cost to serve

The quietest leak is a price that was right once. Freight and packaging inflation never reached the unit economics of a pet brand’s bestsellers in [Bestsellers were priced below their true landed cost](/cases/bestsellers-were-priced-below-their-true-landed-cost/), and resin and energy surcharges rose faster than contracts captured them in [A surcharge lag was never passed into customer contracts](/cases/a-surcharge-lag-was-never-passed-into-customer-contracts/). In services and software the cost is people and support rather than freight: [Retainers were priced below delivery cost](/cases/retainers-were-priced-below-delivery-cost/). The test is the full landed cost of your ten best sellers, shipping, returns and payment fees included, against the price actually paid.

## Pricing and discount findings across the 74 audits

Pricing and discount findings in the published audits, grouped by where the margin went (18 of 74 cases)

| Where the margin went | Cases | Annual figures | One of them |
| --- | --- | --- | --- |
| Codes, promotions and bundles | 4 | $38k to $68k | [A discount habit was training the best customers to wait](/cases/a-discount-habit-was-training-the-best-customers-to-wait/) |
| Free shipping, delivery and extras | 3 | $47k to $58k | [Free shipping was quietly deleting the margin on every third order](/cases/free-shipping-was-quietly-deleting-the-margin-on-every-third/) |
| Plan, deal and channel discounts | 4 | $49k to $92k | [Distributor margin was set once and never revisited](/cases/distributor-margin-was-set-once-and-never-revisited/) |
| Rebates and fee waivers | 2 | $58k to $118k | [Volume rebates were paid on orders that never hit the tier](/cases/volume-rebates-were-paid-on-orders-that-never-hit-the-tier/) |
| Prices below the full cost to serve | 5 | $52k to $79k | [Rising freight never reached the price list](/cases/rising-freight-never-reached-the-price-list/) |

Every figure is a published case’s audited annual value. Across all 74 cases the lower median finding is $58k a year, and the [ranked ledger of findings](/findings/) lists every one. The five [margin leak patterns](/margin-leaks/) page defines mispriced offers and discount habits, with the audits behind each.

## The arithmetic of a discount

A discount costs more than its headline because it comes straight out of the margin, not out of the price. At a 40% [contribution margin](/glossary/contribution-margin/), a 10% discount leaves 30 points of margin on each sale, so you need a third more units just to earn the same contribution. A 20% discount needs twice the units. For a discount d and a contribution margin m, both as shares of the price, the extra volume you need is d divided by (m minus d).

Extra units needed to keep the same contribution after a discount (arithmetic, not audit data)

| Discount | At a 30% margin | At a 40% margin | At a 50% margin |
| --- | --- | --- | --- |
| 5% off | 20% | 14% | 11% |
| 10% off | 50% | 33% | 25% |
| 15% off | 100% | 60% | 43% |
| 20% off | 200% | 100% | 67% |

Revenue based measures miss this entirely. A promotion lifts revenue, and with it ROAS and [MER](/glossary/mer/), in the same weeks that it cuts the contribution underneath both. A discount that pays for itself shows up in contribution per customer, not in the top line. When it does not, the gap is a [margin leak](/glossary/margin-leak/) that grows with every order.

## How to run a pricing audit on your own exports

1. **Pull a year of order lines.** List price, price paid, the discount or code used, the SKU or plan and the channel, one row per line.

2. **Attach the full cost.** Landed cost per SKU, shipping and packaging, payment fees and returns, and for services the delivery hours. Use this year’s costs, not the ones the price was set on.

3. **Compute contribution per order after every discount.** Sort it by SKU, by code and by channel. The losers are usually a short list.

4. **List every standing discount.** Thresholds, plan discounts, dealer terms, rebates, waivers and surcharge clauses, each with the date it was set and the reason. Any reason that no longer holds is a candidate.

5. **Test before you cut.** Withhold a discount from a random share of customers or a set of regions for a few weeks and compare. A discount that brings no extra contribution can go; the method is the same one used for [measuring incrementality](/insights/how-to-measure-incrementality/) in ad spend.

Most of this runs on exports you already have. The [marketing audit checklist](/marketing-audit-checklist/) includes the pricing checks and scores them in your browser.

What it means for your margin

A pricing fix needs no new customers and no new spend: it changes what you keep from the orders you already win. A discount habit and a non incremental campaign are the same leak seen from two sides, money spent on sales you would have made anyway. That is why the audit tests pricing beside paid media, in the same pass.

## Questions about pricing audits

### What is a pricing audit?

A pricing audit checks whether each price, discount, rebate and promotion still covers its full cost and earns the margin it was set to earn. It covers everything given away after the list price, not only the price list. Mispriced offers appeared in 47% of the 74 audits. [The five margin leak patterns →](/margin-leaks/)

### How do I know if discounts are hurting my margin?

Two signs. The share of orders that use a code keeps rising while full price demand falls, and the extra volume a discount brings is smaller than its arithmetic needs: at a 40% contribution margin, a 10% discount needs a third more units to break even. Discount habits appeared in 34% of the audits. [Contribution margin, defined →](/glossary/contribution-margin/)

### How often should prices and discounts be reviewed?

Every time an input moves, such as freight, materials, carrier rates, platform fees or the customer mix, and at least once a year otherwise. Most of the pricing cases in the audits began as sound decisions that nobody revisited after their costs changed. [The marketing audit checklist →](/marketing-audit-checklist/)

[See every finding, ranked by annual value →](/findings/)

## More field notes

[Where SMB marketing budgets leak, and why nobody sees it, Benchmark report](/insights/where-smb-marketing-budgets-leak/) [Your ROAS is lying to you: an attribution field guide, Paid media](/insights/roas-attribution-field-guide/) [How to audit your agency without firing your agency, Working with agencies](/insights/how-to-audit-your-marketing-agency/) [What “good” looks like at $10k to $250k/month spend, Benchmarks](/insights/smb-marketing-spend-benchmarks/) [How to measure incrementality with a simple holdout test, Measurement](/insights/how-to-measure-incrementality/)

[Request an audit →](/book/)
 [See the benchmark](/benchmark/)