Coupons that cost more than they earn: testing promotion profitability

Marketplace dashboards make promotions look successful. Sales rise while the coupon runs, and the report credits the coupon with every sale that used it. The real question is narrower: how many of those sales would not have happened without the discount, and did their margin pay for the discount given to everyone else?

The arithmetic

Every promotion gives away margin on two kinds of sale:

• Baseline sales: customers who would have bought at full price. The discount on these is pure cost.
• Incremental sales: customers who bought because of the promotion. These add margin, reduced by the discount.

The promotion pays only if the margin on incremental sales exceeds the discount on baseline sales plus any promotion fees.

Break-even lift

A quick test before any promotion: the sales lift needed to break even is the discount divided by the contribution remaining after the discount. In the illustration, 2.00 ÷ 4.00 = 50%. A coupon on that product needs to lift sales by more than half just to stand still.

Measuring the real lift

• Compare against a matched baseline: the same weeks of the previous year, or similar products without the promotion.
• Look at the weeks after the promotion; a dip there means sales were pulled forward rather than created.
• Separate the promotion from simultaneous changes in advertising or price.
• Check ranking effects: if a promotion improves organic rank, part of its value appears after it ends.

What to do with the result

Keep promotions where the measured lift clears the break-even lift with margin to spare, usually on products with high contribution and price-sensitive demand. Stop or shrink them where they mainly discount loyal buyers. Write the rule down, so that the next promotion is judged before it runs, not after.