What a discount really costs
Twenty per cent off sounds like it costs twenty per cent of something. It does — twenty per cent of the price. The problem is that it is subtracted entirely from the profit, because your costs do not go down when you drop the price.
The arithmetic
Sell at 100 with a cost of 60. Profit is 40, a 40% margin. Now discount 20%:
- New price: 80
- Cost: still 60
- New profit: 20
The price fell by a fifth. The profit fell by half.
The general rule is simple enough to do in your head:
share of profit lost = discount / margin
A 20% discount on a 40% margin loses 20/40 = half the profit. On a 25% margin it loses 20/25 = four fifths. On a 20% margin it loses all of it.
How much profit each discount leaves you
Find your margin down the left, the discount across the top, and read off how much of the profit survives.
| Margin | −10% | −15% | −20% | −30% |
|---|---|---|---|---|
| 20% | 50% | 25% | 0% | loss |
| 25% | 60% | 40% | 20% | loss |
| 30% | 67% | 50% | 33% | 0% |
| 40% | 75% | 63% | 50% | 25% |
| 50% | 80% | 70% | 60% | 40% |
The row that should worry most online sellers is 25%. After marketplace commission, fulfilment and advertising, that is where a great many products actually sit — and at that level a routine 20% promotion leaves a fifth of the profit standing.
“We will make it up in volume”
Sometimes true. Usually not, and the multiple required is larger than people guess.
volume multiple needed = margin / (margin − discount)
At a 40% margin, a 20% discount needs double the units to earn the same money. At 30%, it needs triple. At 25%, five times.
And those extra units are not free. They consume stock, fulfilment, packaging, customer service and returns handling — all of which scale with volume while your fixed costs stay put. Doubling units to hold profit flat means twice the work for the same result, plus more capital tied up in inventory.
Before running a promotion, put the required multiple next to your realistic expectation of the uplift. If you need triple and you expect fifty per cent more, the promotion is a decision to earn less.
Where the discount actually lands
On a marketplace the discount is worse than it looks, because commission is charged on the price the customer pays. Cut the price and you do save a little commission — but you lose the full discount from a margin that was already net of that commission.
Free delivery is the same conversation in disguise. Absorbing 5 of postage on a 50 item is a 10% discount on the price and, at a 25% margin, forty per cent of the profit. It simply does not appear on the price tag, which is exactly why it gets agreed to more easily than a straight markdown.
What to do instead
- Know your break-even price before the promotion, not after. That is the floor, and it moves every time your advertising cost per unit changes.
- Discount the bundle, not the unit. Three-for-two raises the order value and spreads fixed per-order costs across more units, which a straight markdown does not.
- Discount slow stock deliberately. Clearing inventory that is accruing storage costs is a real saving; discounting your best seller is just giving money away to people who were going to buy anyway.
- Use thresholds. "Free delivery over 60" costs you only on orders that would otherwise have been smaller.
- Check whether you are the cheapest already. Discounting into a position you already hold buys nothing.
Run it on your own numbers
The markup and margin calculator converts between the two and prices to a target margin, so you can see what a discounted price does to the figure that matters. The break-even calculator shows how far volume has to move before a lower price still covers your fixed costs.
Do both before the promotion goes live. Discounts are easy to start and surprisingly hard to withdraw — customers who learn that a product goes on sale every month stop buying it at full price.
Frequently asked questions
Slightly, since commission is a percentage of the price paid. It is a small consolation: on a 15% commission, a 20 discount saves 3 of commission while costing you the other 17 in full.
Above 20% simply to break even on the discounted sale, and closer to 40% for the promotion to still be worth running. Below 25%, routine discounting is not a marketing tactic, it is a slow way of losing money.
Only if the delivery cost is smaller than the discount you would otherwise give. It often converts better for the same money, which is a genuine argument for it — but it is still a discount and belongs in the same calculation.
Slowly, and by changing the offer rather than just the price. Bundles, sizes and variants let you reset the reference price without visibly raising it. Simply putting the old price back usually stalls sales for a while.