Running Discounts at the Register Without Bleeding Margin
A 20% discount needs far more extra sales than you think just to break even. How to design POS promos that grow profit, not just revenue and workload.
Here’s the math almost nobody does before hanging the “20% OFF” banner. Say your gross margin is 40%. A 20% discount cuts that margin to 20% — half of it gone. Which means that just to earn the same profit as before the promo, your sales volume has to double. Not rise 20%. Double.
Most promos never get there. What actually happens: revenue ticks up a little, the register gets busier, and at month’s end profit is thinner than before. That’s the slow bleed — it never feels like losing money, but the numbers say otherwise.
None of this makes discounts evil. A discount is a tool, and tools work fine once you know how to hold them.
Before anything else: know your margin per product
This is the non-negotiable prerequisite. Fifteen percent off a 60%-margin product is relaxed; fifteen percent off a 20%-margin product is selling while burning cash. If your POS records cost prices, the margin report already exists — and discount decisions stop being vibes and become arithmetic.
The rule of thumb is plain: the thinner the margin, the smaller the discount room. Thin-margin products are better promoted through bundles than through direct price cuts.
Discounts that work always buy something
A healthy discount is never free — it trades the price cut for something specific. Formats with a real track record in retail and F&B:
- A spending threshold: “10% off purchases above Rp100,000” — the discount is funded by a bigger basket. Set the threshold slightly above your average transaction value (your POS report has that number).
- Bundles: the rice-chicken-iced-tea set at a package price. The fat-margin item (always the drink) subsidizes the cut, and the customer feels the abundance.
- Quiet-hour discounts: happy hour from 2 to 5 p.m. An empty chair can’t be stored — filling it at a thinner margin beats leaving it empty. Never run it during rush hours; that’s discounting people who were buying anyway.
- Next-visit vouchers: a coupon valid next week trades the discount for a return visit — and whoever doesn’t return never collects the cut.
Notice the pattern: each format purchases something — bigger baskets, filled dead hours, repeat visits. A discount that buys nothing (“20% off everything, because it’s been slow”) merely transfers your money to whoever happened to walk past.
The psychological line: don’t train customers to wait
Promos that run too often create a new problem: customers learn that the normal price is fiction. Look at fashion retail chains on permanent “sale” — nobody will ever pay full price there again.
The fix: every promo needs a start, an end, and a story you can tell (payday week, store anniversary, season). When it ends, it truly ends. The gap between promos is what makes the next one feel special. And for regulars, a points or stamp program usually beats open discounts long-term — the value flows to people who genuinely keep coming back, along the lines of our WhatsApp loyalty program guide.
Run it through the system, not through the cashier
The underrated half: execution. A promo whose rules exist only as verbal instructions (“give 10% off if anyone asks”) will leak — miscalculated, applied after the end date, or quietly extended to the cashier’s friends. The abuse patterns are exactly the ones from detecting cashier fraud.
Promo rules belong inside the POS: active dates enforced automatically, minimum-spend conditions checked automatically, and any manual discount outside the rules requiring supervisor approval. Beyond tidiness, this makes evaluation honest — every price cut is logged under its promo name.
Evaluation: one question only
When the promo ends, open the report and answer a single question: was gross profit during the promo period higher or lower than a comparable normal period? Not revenue — gross profit. Growing revenue is trivial; just deepen the discount. Growing profit is proof the promo worked.
While you’re there, check the side effects: did the week after the promo crater (customers merely pulled purchases forward)? Did non-promo items get lifted along the way? Those two signals separate promos that build a business from promos that borrow from its future.
Start with the lowest-risk version: one promo, one month, with a minimum spend, on a product whose margin you know cold. Measure gross profit, then decide whether it earns a second run. A measured discount is a strategy. An unmeasured one is just a lower price.