A discount feels like a marketing decision and behaves like a financial one. The banner goes up in an afternoon; the cost of it is paid quietly, per unit, for as long as the promotion runs, in the only money the business actually keeps — its margin. Owners who would never sign a contract without reading it routinely sign promotions without computing them, because the price reduction is visible and immediate while the margin reduction is neither.
The good news is that a promotion needs only three calculations to be judged honestly. What does the discounted price actually become, once stacking and tax are handled correctly? What does the discount do to the contribution each sale makes? And — the question almost never asked — how much extra volume must the promotion generate for the business to end up no worse off than selling fewer units at full price?
This article works through all three, with numbers small enough to check on a napkin, and finishes with the promotion designs that experienced sellers prefer precisely because their costs are bounded and known in advance.
The stacking trap: percentages multiply, they do not add
“25% off, plus an extra 10% this weekend” reads as 35% off. It is not. The second discount applies to the price the first one left behind: 100 falls to 75 under the first, and 10% of 75 is 7.50, leaving 67.50 — a total reduction of 32.5%. The customer receives less than the banner implies; the seller who budgeted for 35% has mispriced the promotion in the opposite direction. Stacking always under-delivers compared with addition, because each percentage acts on a shrinking base.
Tax sharpens the point. In most systems the tax is charged on the amount actually paid, after discount — so a discounted sale also reduces the tax collected on it, and the customer's total saving is slightly larger than the shelf-price saving. Sellers modelling a promotion on the pre-tax shelf price alone will misstate both their margin and the customer's bill. The sequence is always: price, discounts in order, then tax on the remainder. Any model that applies tax earlier in the chain is wrong in a way that matters at volume.
What a discount does to contribution
Take a product priced at 100 with a unit cost of 60: contribution 40 per sale, a 40% margin. Cut the price 20% and the new price is 80 — but the cost did not attend the meeting, so the contribution falls to 20. The price dropped by a fifth; the contribution halved. This asymmetry is the central fact of discounting: because the discount comes entirely out of the profit slice, modest price cuts amputate contribution at two, three or four times their rate.
Push the cut to 40% and the price is 60 — contribution zero. The business now sells at cost, converting stock into revenue and nothing else, while still paying the per-sale costs hidden in the 60 or outside it. Discounts feel linear from the customer's chair. From the seller's chair they are leveraged, and the leverage grows as the starting margin thins: for a product on a 15% margin, a 15% discount is not a trim, it is the entire profit plus a small donation.
The volume question: how many more must you sell?
A promotion pays for itself only in volume. The required increase follows from the contributions: you need (old contribution ÷ new contribution) times as many sales to bank the same gross profit. For the 100-to-80 cut above, that is 40 ÷ 20 = twice the sales — a 100% volume increase — merely to stand still, before counting the promotion's own costs: the advertising, the extra staff hours, the returns that sales events bring with them.
Write that required increase down before the promotion launches and judge it as a forecast. Has this shop ever doubled volume on this line in a week? Does the stock exist? Can fulfilment cope without errors that generate refunds? If the honest answer is no, the promotion is a planned loss — sometimes still worth running, for stock clearance or customer acquisition, but then it should be entered in the books as a marketing expense with a ceiling, not discovered in them as a margin collapse. Clearance has its own logic, by the way: an item sold at a small loss today may beat a warehouse cost and a write-off tomorrow — but that logic compares against holding the stock, not against the full-price fantasy.
Designs that keep the cost bounded
The promotions professionals repeat share a feature: their cost is capped by design. A fixed amount off above a threshold (“10 off orders over 120”) spends at most 10 per qualifying order and lifts basket size. A free add-on of known cost — the item costs you 6, delights like 15 — gives perceived value at wholesale cost. A time-boxed discount on a single line contains the blast radius. A loyalty reward earned on full-price purchases discounts the next visit rather than this one, protecting today's margin entirely.
What they avoid is the open-ended percentage off everything, whose cost scales with your best customers' biggest baskets. And whatever the design, two closing habits separate promotion from leakage: a published end date that is actually honoured, so customers learn that waiting you out does not work; and a post-promotion review comparing contribution earned against the same period at full price, so the next decision uses evidence instead of the memory of a busy till. Busy is not the metric. Contribution is.
Key takeaways
- Stacked discounts multiply: 25% then 10% is 32.5% off, not 35% — and tax is charged after discounting, on the reduced price.
- A discount comes entirely out of contribution: a 20% price cut on a 40% margin halves the profit per sale.
- Required volume increase = old contribution ÷ new contribution. Double-check the promotion against that figure before launching.
- Prefer bounded designs: amount-off thresholds, free add-ons at known cost, single-line or time-boxed offers.
- Review every promotion afterwards by contribution earned, not by how busy the shop felt.
Mistakes to avoid
- Advertising stacked percentages as their sum — disappointing customers and misstating your own costs simultaneously.
- Judging a promotion by revenue or footfall while contribution per sale quietly halves.
- Running percentage-off-everything offers whose cost is uncapped and lands hardest on your largest orders.
- Extending “one-week” sales for months, training customers never to pay full price.
Questions readers ask
How deep a discount can a 30% margin product survive?
Directly: contribution is 30 per 100 of price, so a 10% discount leaves 20 (needs 1.5× volume), 20% leaves 10 (needs 3× volume), and 30% leaves nothing. Depth is survivable only if the volume genuinely follows — compute the required multiple first.
Is it better to discount the price or add something free?
Usually the add-on: its cost to you is the wholesale cost of the item, while its perceived value is the retail price. A discount costs its full face value, in margin, on every unit sold.
Should clearance stock be judged the same way?
No. Once stock is bought, its cost is sunk; the comparison is between selling cheaply now and the likely outcome of holding it — storage, deterioration, eventual write-off. Clearance arithmetic starts from “what is this worth if unsold?”, not from the original margin.
Put it into numbers
Model your next promotion properly: original price, discount, any stacked offer and the tax, in the right order — then read the true saving and decide if the volume can follow.
Open the Discount Calculator