Price waterfall: from list price to pocket price, and where the margin leaks
A price waterfall is a chart that starts at the list price and subtracts every discount, rebate and cost of the sale. What remains is the pocket price: the revenue you keep from the transaction. This guide builds one with example numbers, shows how to calculate price realization, and explains how an approval on each quote stops the leaks.
What a price waterfall shows
The idea comes from McKinsey’s pricing work. Michael Marn and Robert Rosiello described this approach to transaction pricing in Harvard Business Review in 1992. A 2003 McKinsey Quarterly article, The power of pricing, revisited the pocket price waterfall. It describes the tool as one that “shows how much revenue companies really keep from each of their transactions.”
The waterfall has three price levels:
- List price. The published or standard price before any discount.
- Invoice price. The list price minus the discounts printed on the invoice, such as a distributor discount or an order-size discount.
- Pocket price. The invoice price minus everything that happens off the invoice. That includes cash discounts, rebates, marketing allowances, freight the seller pays and the cost of waiting to be paid.
The off-invoice items are the ones nobody sees on a single document. They sit in rebate agreements, freight bills and the receivables ledger, so they do not appear in the margin a salesperson sees on the invoice. McKinsey adds the pocket margin for companies whose products or cost to serve vary by customer: the pocket price minus direct product costs and the costs incurred to serve that account.
A price waterfall example
Take an HVAC unit with a list price of $10,000.00, sold through distributors. The waterfall below averages one distributor’s orders over a year. McKinsey notes that a waterfall is often first built as an average of all transactions. The numbers are made up to show the method.
| Waterfall step | How it is worked out | Per unit |
|---|---|---|
| List price | Published price | $10,000.00 |
| Standard distributor discount | 25% of list, on the invoice | −$2,500.00 |
| Order-size discount | 3% of list, on the invoice | −$300.00 |
| Discretionary discounts | Average per unit, given to win orders | −$400.00 |
| Invoice price | 68.0% of list | $6,800.00 |
| Cash discounts taken | 2% on about half the invoices, so 1% of invoice on average | −$68.00 |
| Annual volume rebate | 3% of invoice | −$204.00 |
| Co-op marketing allowance | 1% of invoice | −$68.00 |
| Freight paid by the seller | Average freight bill per unit | −$310.00 |
| Cost of carrying the receivable | 45 days on average at 8% a year | −$67.07 |
| Pocket price | 60.8% of list | $6,082.93 |
The invoice says the distributor pays $6,800.00. The plant keeps $6,082.93, and the five off-invoice items take 7.2% of list between them. With a unit cost of $4,700.00, the gross margin on the invoice price looks like 30.9%. The margin on the pocket price is 22.7%.
No single item looks large. The freight is one bill among hundreds, and the rebate is paid once a year. The waterfall puts them in one place, per unit, and names each part of the gap between 30.9% and 22.7%.
How to build your own price waterfall
- Pick one product family and the last twelve months of invoice lines for it.
- Group the on-invoice discounts by type, and keep list price on every line.
- Add credit notes, rebates and co-op payments, each assigned to the customer and period it belongs to.
- Add freight from the carrier bills, by shipment, and the days each invoice took to be paid.
- Draw the average waterfall first, then one per customer, and sort the customers by pocket price.
The first draft takes the longest, because rebates and freight sit in different systems from the invoices. Keep the matching rules written down, so next quarter’s waterfall uses the same ones.
Price realization
One way to measure price realization is the share of list price that turns into revenue. In the example, the pocket price is 60.8% of the list price. Track it per customer, per product family and per salesperson, because the average hides the spread.
Manufacturers also report price realization as the change in their net prices from one period to the next. Caterpillar’s second-quarter 2026 results attribute $595 million of its sales increase to favorable price realization, beside $3.1 billion from higher sales volume. In the example, if last year’s pocket price for the same unit was $5,850.00, this year’s $6,082.93 is 4.0% higher.
Small changes in realized price move profit more than they seem to. McKinsey’s 2003 article worked through the average income statement of an S&P 1500 company. A 1% price rise at stable volume would raise operating profit by 8%, it found.
See where your discounts leave the price
Send Derik the workbook your team quotes from. He sends back a free map of its rules, including every discount it allows, and its gaps, such as prices typed by hand with no source.
Start a conversationWhere margin leaks
Margin leakage, or price leakage, is revenue lost between the price you meant to charge and the pocket price. McKinsey’s example is a lighting supplier whose on-invoice discounts put average invoice prices 32.8% below list. Off-invoice items took another 16.3 percentage points, leaving a pocket price of about half the list price.
The same article found a wide pocket price band at that supplier. Some bulbs sold at less than 30% of list, and others at 90% or more. Look for leaks in these places:
- Discretionary discounts. Discounts given to win an order, with no record of who approved them or why.
- Order-size discounts on small orders. A discount tier applied to orders that never reached the tier.
- Cash discounts taken late. A customer deducts the prompt-payment discount but pays after the discount period.
- Rebates on volume never reached. A rebate rate set for a volume forecast the customer missed.
- Freight the seller absorbs. Prepaid freight on orders below the free-freight threshold, or expedited shipping nobody billed.
- Payment terms that stretch. Every extra day the receivable stays open costs interest, as the carrying cost in the example shows.
Cash discounts deserve a closer look. OpenStax’s financial accounting text explains terms such as 2/10, n/30: a 2% discount for paying within 10 days, with the full amount due in 30. Giving up 2% to be paid 20 days sooner costs the seller about 37.2% a year. If a customer takes the discount and still pays late, the seller pays that rate for nothing.
Discount policy, discount schedule and discount matrix
A discount policy says who may give which discount, and who approves anything beyond it. A discount schedule lists the standard discounts, usually by customer type or order size. A discount matrix puts the two dimensions in one table, like this example:
| Customer type | Order under $5,000 | $5,000 to $25,000 | Over $25,000 |
|---|---|---|---|
| Stocking distributor | 25% | 27% | 30% |
| Non-stocking distributor | 20% | 22% | 25% |
| Equipment maker buying direct | 15% | 18% | 20% |
In this example policy, a discount up to 3 points above the matrix needs the sales manager’s approval. Anything more needs the owner. The numbers are made up, but the structure is the point: every discount is either in the matrix or has a name and a reason attached.
OpenStax separates a trade discount, negotiated before the purchase, from a cash discount, earned by paying early. Keep the two in separate columns of the waterfall, because different people control them.
Stop the leak with a named approver on each quote
A waterfall drawn once a year explains last year. To change next year, the same arithmetic has to run on every quote before it goes out.
The approver needs the whole waterfall for the quote, because the discount percentage alone hides the rest. A 3-point discount on a customer who takes every cash discount and prepaid freight costs more than the same discount on one who does not. Human in the loop covers how to keep an approval step from turning into a rubber stamp.
Where AI does the work
The hard part of a price waterfall is the data. The off-invoice items live in documents and ledgers that do not share a customer key. AI does the reading and matching, and a named person approves any change to a price or a policy.
- Building the waterfall per transaction. The software joins invoice lines with credit notes, rebate agreements, freight bills and payment dates, so each order gets its own waterfall.
- Reading the agreements. Language models read rebate and co-op agreements written as PDFs and turn their tiers and rates into rules a person confirms.
- Finding the band. It sorts customers by pocket price and flags the ones far below others of similar size, with the items that put them there.
- Checking each quote. Before a quote goes out, it estimates the pocket margin and routes exceptions to the named approver.
Margin analysis covers margin by product, customer and mix, and price optimization covers setting prices from cost, demand and quote history.
Before you change discounts
McKinsey’s article points out that a wide pocket price band can reflect real differences between customers. Different prices to competing buyers can also carry legal risk when you sell into the United States. The Federal Trade Commission explains that charging competing buyers different prices for the same commodity may violate the Robinson-Patman Act. It names two main defenses. One is a price difference justified by different costs of manufacture, sale or delivery. The other is a concession made in good faith to meet a competitor’s price. Keep the reason with every exception, and ask counsel before you change discount structures for competing distributors.
How ThriveAI builds your price waterfall
ThriveAI’s margin and pricing offer builds margin by product, customer and channel from your invoice lines, credit notes, rebates and freight. Derik connects read-only to your ERP and accounting system and ties the result to your income statement. You see margin by customer after rebates, freight and credit notes, which comes close to the pocket margin this page describes. You also get the prices that fall below your margin floor, each with a drafted new price. You approve, change or reject each one. The first stage runs four weeks on your own workspace. To check quotes before they go out, the Excel builder to CPQ app offer adds the same check to every quote, with a named person approving.