What Discount Depth Does to Contribution Margin
A 30 percent discount does not cost you 30 percent. On a 40 percent margin product it takes three quarters of your contribution, and the volume needed to cover that is usually not available.
- On a 40 percent contribution margin, a 20 percent discount needs double the units to break even and a 30 percent discount needs four times.
- Break-even volume multiple equals contribution margin percent divided by contribution margin percent minus discount percent.
- At a 30 percent contribution margin, a 30 percent discount has no break-even at all, because contribution goes to zero.
- Promo cohorts typically repeat at a lower rate than full-price cohorts, so the effective acquisition cost is higher than the promo sheet shows.
Discount decisions in most Indian D2C teams get made in a chat thread on a Thursday afternoon. Somebody suggests 25 percent for the weekend, somebody else says make it 30, and the banner goes live. Nobody runs the arithmetic, which takes about ninety seconds and usually changes the answer.
The arithmetic nobody runs
A discount does not cost you the discount percentage. It costs you a share of your contribution, and that share is far larger than the number on the banner.
Take a product at Rs 1,000 with Rs 600 of variable cost: landed COGS, shipping, payment fees and the returns provision. Contribution is Rs 400, or 40 percent. Now cut the price.
- 10 percent off. Price Rs 900, contribution Rs 300. You need 1.33 times the units to stand still, so plus 33 percent volume.
- 20 percent off. Price Rs 800, contribution Rs 200. You need 2 times the units, so plus 100 percent volume.
- 30 percent off. Price Rs 700, contribution Rs 100. You need 4 times the units, so plus 300 percent volume.
The formula is one line. Break-even volume multiple equals contribution margin percent divided by contribution margin percent minus discount percent, both measured against the original selling price.
At a 30 percent contribution margin the picture worsens fast. A 10 percent cut needs plus 50 percent volume. A 20 percent cut needs plus 200 percent. A 30 percent cut has no break-even at all, because contribution is zero and every extra unit is free labour for the customer.
Two adjustments make it harsher in practice. Marketplace and quick commerce commissions are charged on the discounted price, which helps slightly, but shipping, packaging and payment gateway costs are close to fixed per order, so they eat a bigger share of a smaller ticket. And promotional volume returns at a higher rate than base volume in most categories, so the returns provision you used at full price is understated.
Why the volume is usually not there
Doubling units on a 20 percent cut assumes your category has that much unused price elasticity. Occasionally it does. Usually the lift comes from three places, and only one of them is worth having.
- Pull-forward. Existing customers who were going to buy next month buy now. No incremental revenue, just discounted revenue moved earlier. It reads as a strong weekend followed by a dead fortnight.
- Deal seekers. New buyers acquired at a price they will expect again. Their repeat rate sits well below your organic cohort, so the discount buys you a worse customer at a higher effective cost.
- Genuine trial. A buyer who would not have tried at full price and who then repeats at normal rates. This is the only cohort the discount was actually for.
Measure the split before the next promotion. Tag the promo cohort and pull its 60-day and 90-day repeat rate against a matched full-price cohort. If promo buyers repeat at half the rate, your effective acquisition cost is double what the spreadsheet claims, and the promotion was an expensive acquisition channel rather than a revenue event.
What depth does to price expectation
Price is a signal, and customers learn the pattern faster than teams expect. If your site runs 25 percent off in three weeks out of four, your list price is fiction and 25 percent off is your price. The symptom is a collapse in full-price conversion during the gaps, and rising cart abandonment as customers wait for the next banner.
Depth also resets what a discount has to be next time. Once a customer has bought at 30 percent off, 15 percent reads as a poor deal. You have permanently raised the floor on your own promotional spend in exchange for one good weekend.
Frequency matters as much as depth. Two 30 percent events a year read as events. Twelve of them read as pricing.
Designing a promotion that does not reprice you
The objective is to move a specific behaviour without changing the reference price of your hero SKU. Structures that manage it:
- Value adds instead of price cuts. A free sachet, a sample of an adjacent SKU, a travel size. You give up variable cost, not contribution on the full ticket, and the price on the product page never moves.
- Threshold offers. Rs 200 off above Rs 1,499 raises basket value instead of lowering unit price, and the effective discount as a share of basket is lower than it feels to the customer.
- Bundles at a new price point. A three-pack carries its own reference price and its own SKU. The single unit stays clean.
- Depth on tail SKUs, not heroes. Clear slow movers hard and protect the SKU that customers price-check across channels.
- Time-boxed and named events. A dated window with a stated end trains customers to expect scarcity. An always-on banner trains the opposite.
Set the guardrails before the season, not during it: maximum depth by SKU tier, maximum promotional days per quarter per SKU, and a floor contribution per order below which the offer switches off automatically. Write them down, and give one named person the authority to say no on a Thursday afternoon.