Xero’s data on Australian small business notes something uncomfortable: sales events now “drive volume but not always profitability.”
Most retailers already suspect this. Very few have run the number that proves it.
The break-even arithmetic
Required volume increase = discount ÷ (gross margin − discount)
| Gross margin | 10% off | 20% off | 30% off |
|---|---|---|---|
| 30% | +50% | +200% | Impossible |
| 40% | +33% | +100% | +300% |
| 50% | +25% | +67% | +150% |
| 60% | +20% | +50% | +100% |
At a 40% gross margin, a 20% discount requires you to double units just to stand still on gross profit.
Did your last sale double units? Almost certainly not — which means it moved stock and destroyed margin.
Cannibalisation makes it worse
The volume you do get is not all incremental. A meaningful share comes from customers who would have bought at full price and simply waited.
Train customers to expect a sale every six weeks and they will stop buying in between. That is not a promotion strategy; it is a permanent price reduction with extra admin.
The test: if your full-price weeks are getting quieter, your promotions are eating them.
When markdown is correct
Discounting is not a margin strategy. It is an inventory strategy, and it has legitimate uses:
Clearing ageing stock. Stock unsold at 180 days is not an asset — it is cash already spent, occupying space and financing capacity. With interest costs up 36.3% since March 2020, holding it has a real cost. Clear it, book the loss, free the capital.
Acquiring genuinely new customers — where you can measure that they are new, and you know your lifetime value.
Defending share against a specific competitive event — time-boxed, with an end date decided in advance.
Each of these is a deliberate purchase of something with a known cost. “It felt busy” is not.
The three-question test
Before any promotion, answer in writing:
1. What volume increase does this need to break even? Use the table.
2. What proportion will be cannibalised from full-price sales you would have made anyway?
3. What is the actual objective — clearing stock, acquiring customers, or defending share? Three different goals, three different success measures.
If you cannot answer all three, it is a hope, not a plan.
Better tools than discounting
Bundle instead of discount. Two items at a combined price protects the perception of individual value and lifts transaction size.
Add value instead of cutting price. Free delivery, extended warranty, a service inclusion. Costs you less than the equivalent discount and does not reset the reference price.
Segment the offer. A discount to lapsed customers only, or to a list, does not train your whole base to wait.
Time-box hard. A sale with a genuine end date creates urgency. A permanent “sale” section is just your price list.
Markdown ladder for ageing stock. 15% at 90 days, 30% at 120, 50% at 180. Systematic, planned, and it stops the annual panic clearance.
Measure what actually happened
After every promotion, four numbers:
- Units sold versus the equivalent non-promotional period
- Gross profit dollars — not revenue, not units. Dollars.
- Full-price sales in the weeks before and after — this is where cannibalisation shows
- New versus existing customers, if the objective was acquisition
Most retailers measure revenue and feel good. Gross profit dollars frequently tell a different story, and it is the only one that pays wages.
Where to go next
- Retail Margin Compression\n- Gross Margin by Service Line\n- How to Raise Prices Without Losing Customers\n- Pricing for Profit
Sources: Xero Small Business Insights Australia 2026; RBA Statement on Monetary Policy May 2026; AMP Bank GO Small Business Cost Pressure Index 2026. General information, not financial advice.