How to use Margin-Aware Discounting
A practical framework for sizing promotions against your gross margin floor — including the math, depth ceilings by category, and how to run a promo calendar that doesn't quietly bankrupt the P&L.
Margin-Aware Discounting
A pricing discipline that sizes promotional depth against gross margin, so discounts never push effective margin below a defined floor.
Margin-aware discounting is the practice of setting promo depth as a function of an item's gross margin, not its sticker price. Every percentage point of discount comes directly out of margin — so a 20% sitewide promo on a 55% gross-margin item drops you to 44% effective margin, not 35%.
The framework gives operators a defined margin floor (the lowest effective margin you're willing to accept after discount, shipping, payment fees and returns), a calendar of allowed promo windows, and depth ceilings per product tier. It replaces the reflex 'match the competitor's 30% off' with a number you can actually defend in the next P&L review.
Most stores discover their margin floor the wrong way — after a Black Friday run where the headline revenue looked great and the gross profit line looked sick. The discount looked like 25% off the customer-facing price. On the P&L, it was 25 points of margin gone.
This guide gives you the four pieces you need: the math that links discount depth to effective margin, the calendar rules that stop discounts compounding, the depth ceilings by category, and an operating playbook for sign-off. It sits inside the broader topic of gross margin levers — pricing, COGS and mix — and treats promo depth as the lever you control on the shortest time horizon.
The math: why a 20% discount is not a 20% problem
Effective margin after discount is straightforward arithmetic, but it's the arithmetic that most merchandising calendars skip. If your item sells at €100 with €45 COGS, your gross margin is 55%. Apply a 20% discount and the customer pays €80, leaving €35 of gross profit — a 43.75% effective margin.
The reason the drop feels disproportionate: COGS doesn't move when the price does. Every discount euro is a margin euro. On a 55% gross-margin item, a 20% price cut is an 11.25-point margin cut. On a 35% gross-margin item, the same 20% discount halves your margin to 18.75%.
Then layer the variable costs the marketing team forgets: payment processing (1.5-2.5%), pick-and-pack (€1.50-€3 per order), inbound shipping amortisation, and the return rate for the SKU. On apparel with a 30% return rate, your contribution margin on the discounted unit is what you keep after one in three goes back.
The compounding trap
Stacked promos break the math. A 20% sitewide + a 10% email code + free shipping (worth ~5% on a €80 order) is not a 35% discount — it's a margin event that can land you below contribution-positive on lower-tier items. Always model stacks on the worst-case SKU, not the average.
Promo-calendar discipline
Calendar discipline is the second half of the framework. The math tells you how deep a single promo can go; the calendar tells you how often you can run one before the discount becomes the price. Customers learn your cadence within two cycles — if every email has a code, the full-price baseline disappears.
A defensible cadence for most online retail brands is 4-6 promo windows per year: a Q1 clearance, a mid-year refresh, BFCM, and one or two category-specific moments. Outside those windows, the only acceptable codes are welcome series (first order), winback (90+ days inactive) and VIP. Each has a documented depth ceiling and an expiry.
Effective margin after discount, by starting gross margin
70% GM (beauty, accessories)
55% GM (apparel)
35% GM (electronics, supplements)
The curves diverge fast. A 30% sitewide that's painful on apparel is catastrophic on a 35%-margin electronics SKU — you cross into negative contribution before you even count returns. This is why a single sitewide depth across mixed-margin catalogues is almost always wrong; depth needs to be tiered.
Depth ceilings by category and margin tier
A depth ceiling is the maximum discount a SKU or collection is allowed to take in any single promo, including stacked codes. Set it once per pricing review, encode it in your promo tool, and refuse exceptions outside of end-of-life clearance. The ceiling protects the floor.
Use the table below as a starting point. The 'effective margin floor' column is what you'll land at if you go to the ceiling — anything below that target means you re-tier the SKU or pull it from the promo. For end-of-life and overstock, you can break the ceiling on a per-SKU basis, but only with a sell-through date and a written cap.
Suggested promo depth ceilings by gross-margin tier
| Margin tier | Typical categories | Max single-promo depth | Max stacked depth | Effective margin floor |
|---|---|---|---|---|
| 70%+ GM | Beauty, fragrance, accessories, jewellery | 30% | 35% | ~54% |
| 55-70% GM | Premium apparel, home textiles, footwear | 25% | 30% | ~40% |
| 40-55% GM | Mid-market apparel, bags, lifestyle | 20% | 25% | ~30% |
| 30-40% GM | Supplements, consumables, basics | 15% | 18% | ~22% |
| <30% GM | Electronics, marketplace-priced | 10% | 12% | ~20% |
Two adjustments before you adopt these numbers. First, subtract your blended return rate from the margin floor for any category with returns above 15% — apparel teams routinely overestimate post-promo margin by 8-12 points because they forget the reverse logistics line. Second, exclude shipping subsidies from the discount ceiling and count them separately; free shipping at AOV thresholds is often the most efficient lever and shouldn't get crowded out by a percent-off code.
Operating playbook: who signs off, what gets tested
The playbook turns the framework into a routine. Every promo proposal carries a one-page brief: starting GM by collection, proposed depth, modelled effective margin floor (worst-case SKU and blended), expected uplift versus the baseline week, and a no-go threshold. Finance signs off if effective margin floor is met; merchandising owns the sell-through; CRO owns the on-site mechanic.
What to test inside the framework: gifted thresholds versus percent-off (often margin-positive because the gift is at COGS, not retail), tiered spend-and-save ladders that pull AOV up, bundle pricing that hides the unit discount, and free-shipping thresholds set just above current AOV. Run these as proper A/B tests against the discount baseline — most brands find at least one mechanic that beats a flat percent-off on contribution margin.
The one number to report
Stop reporting promo performance as revenue uplift versus a non-promo week. Report it as incremental contribution margin: (promo-week contribution margin) − (baseline-week contribution margin × seasonal index). If the incremental number is negative, the promo lost you money even if revenue was up.
Frequently asked questions
Start from your gross margin, subtract your minimum acceptable effective margin (the floor), and the difference is your maximum depth. On a 55% GM item with a 40% floor, you can afford roughly 27% off before variable costs. Subtract another 3-5 points for payment fees and returns.
Because COGS doesn't discount with the price. Every discount euro comes out of gross profit, not revenue proportionally. A 20% price cut on a 55% margin item is an 11.25-point margin cut — it scales with your starting margin, not with the discount headline.
Sitewide only works if your catalogue has a tight margin range. Mixed catalogues need tiered depth — a flat 25% off is fine on 65% GM beauty and ruinous on 35% GM electronics. Most stores above €3M revenue should move to collection-level promo rules.
Free shipping is usually cheaper than a 10% code at the same AOV, because the cost is a flat number rather than a percentage of the basket. Setting a free-shipping threshold just above current AOV often beats a percent-off on contribution margin while also lifting basket size.
Often yes — the gift costs you COGS (perhaps €4-€8) rather than a percentage of retail (€20-€30). The trade-off is inventory exposure on the gift SKU and slower perceived value than a clear percent-off in performance creative.
Most online retail brands hold a defensible full-price baseline at 4-6 promo windows per year. Above 8 windows, repeat customers start waiting for the next code and your full-price weeks erode. Track 'share of orders at full price' as a guardrail metric.
Yes — particularly in apparel. A 25% return rate means you keep three of every four discounted units, and the returned unit still cost you reverse shipping plus restocking. Subtract your category return rate from the effective margin floor when sizing depth.
Don't reflexively match. Compare your gross margin to theirs (often visible in their reported numbers if public, or estimable from category norms). If they have 70% GM and you have 45%, matching their 30% off puts you at a 21% margin and them at 58%. Compete on bundles, exclusivity or shipping speed instead.
Incremental contribution margin versus a seasonally adjusted baseline week. Revenue uplift alone hides margin destruction. The full report: incremental units, incremental contribution margin, new-customer share of orders, and full-price recovery in the two weeks after the promo ends.
End-of-life clearance, overstock with carrying costs, and seasonal SKUs past their sell-by window. In each case the alternative — writing off the inventory or holding it through another season — is worse than the margin hit. Document the cap, the sell-through date and the units; don't let one-offs become precedent.
Test ideas before you ship them
Run unlimited A/B tests, attach hypotheses to outcomes, and build a searchable archive of what works — and what doesn't.