Discounts are judged against revenue and paid out of contribution, and that single mismatch is why most promotional calendars are more expensive than the people running them believe. A code takes its percentage off the price, but the goods, the parcel and the fixed part of the payment fee cost exactly the same on a discounted order, so the whole discount comes out of the thin slice the order was actually keeping. On typical economics a ten percent code removes about a quarter of that slice, which means the promotion has to sell more than a third more units before the month has gained anything at all.
- Ten percent off the price is roughly twenty-five percent off contribution, and the required volume lift follows from that rather than from the depth.
- On the example store a 10% code needs about 35,6% more orders just to keep total contribution level.
- Scope beats depth: a shallow code aimed at the right orders usually earns where a deep sitewide one does not.
- Every promotion needs a floor: below the price that covers goods, parcel and fee, volume makes things worse rather than better.
The leverage, stated once
When roughly a third of the net price survives the variable costs, every point of discount comes out of that third rather than out of the whole. So the required lift grows much faster than the discount depth does: ten percent needs about a third more orders, twenty percent needs to roughly double them, and thirty percent is usually beyond anything a promotion can deliver. A discount impact calculator does the division for your own margins in a minute, and the number is almost always larger than the room expects.
| Depth | Contribution left | Orders needed to stand still |
|---|---|---|
| No code | €15,35 | baseline |
| 10% | €11,32 | about 36% more |
| 15% | €9,30 | about 65% more |
| 20% | €7,29 | about 111% more |
Scope is cheaper than depth
A sitewide code hands its discount to every customer already in the checkout, and that share buys no additional volume whatsoever. Scoping the same promotion, to slow stock, to first orders, to a category, to a minimum basket, keeps most of the effect and removes most of the giveaway, which is why a well-aimed ten percent frequently beats a broad twenty.
The same logic applies to how the discount is delivered. Reducing a price rather than issuing a code produces price reductions instead, which look identical to the customer and disappear from most discount reports, so a store that discounts both ways and only tracks one is systematically underestimating its own generosity. Worse, a compare-at price left in place after a promotion keeps producing that reduction forever.
Depth needs a floor underneath it
Before any promotion runs, work out the price at which the order stops covering its own costs. That pricing floor is not a target, it is the line under which a sale makes the store poorer regardless of how many units move, and knowing it turns an argument about how deep to go into a bounded decision.
Clearing dead stock below the floor is a legitimate exception, because the alternative is a write-down and continued storage. The distinction is that it is a decision about inventory rather than a marketing tactic, and it should be scoped, dated and ended rather than becoming the store's habitual price.
Judging a promotion honestly
- Write down the volume the code owes you before it runs, using your own contribution rather than a rule of thumb.
- Compare the promotion period against a comparable one, not against the week before, which is usually depressed by anticipation.
- Count contribution, not revenue. A promotion that lifted revenue and lowered contribution is a loss with good publicity.
- Check what share of the discount went to customers who were already buying, which is the cheapest thing to fix.
- Record the effective reduction across everything sold, not the headline percentage on the banner.
Codes that leak
Three leaks account for most of the difference between what a promotion was meant to cost and what it did. Codes that outlive their campaign because nobody set an expiry, and then circulate on coupon sites forever. Codes that stack with an existing sale price, so a fifteen percent promotion lands as thirty on the products that could least afford it. And codes shared publicly that were built for one segment, which is the same giveaway with none of the targeting.
All three are configuration rather than strategy, which is what makes them worth an hour: an expiry date, a stacking rule and a usage limit remove most of the leakage without changing a single thing about the offer the campaign actually intended.
The uncomfortable finding
Stores that run this arithmetic for the first time usually discover two things at once: their standard promotion has never paid for itself, and one smaller promotion they barely think about earns quietly every time it runs. Neither fact is visible in a revenue report, both are obvious in a contribution one, and the difference between them is generally worth more over a year than any pricing change the same store is willing to make.