Field note · Pricing

Coupon Strategy Without Destroying Contribution Margin

A coupon can increase orders and revenue while reducing the money available to pay overhead. The missing comparison is contribution per eligible visitor. This worked analysis prices a $5 coupon on a $50 product, allows for lower percentage fees, and calculates the conversion rate that would preserve the original contribution. The same assumptions power the editable model and downloadable sensitivity table below.

Build a baseline that includes the cost of traffic

Consider one product with a $50 price, $25 landed goods cost and $5 fulfillment cost. A hypothetical variable fee is 8% of the discounted merchandise price. This is a model input, not a quoted rate for any platform. The baseline receives 1,000 eligible sessions at $0.25 per session, and 3% of sessions become orders. Traffic costs $250 whether 30 or 38 visitors buy.

Without a coupon, contribution before traffic is $50 − $25 − $5 − $4 = $16 per order. Thirty orders generate $480 before traffic and $230 after the $250 traffic bill. Fixed overhead and tax remain outside this example. So do any change in return behavior, basket size or repeat purchases; those must be evaluated separately rather than credited to an unproven promotion.

A 10% price discount needs 40.35% more orders here

The seller gives away $5, but saves $0.40 in percentage fees, so contribution falls by $4.60. Simply subtracting the coupon from the original contribution would overstate the required lift. Conversely, ignoring the lost contribution because the platform fee got smaller would understate it.

The conversion rate must rise by 1.2105 percentage points, or 40.35% relative to its baseline. At 1,000 sessions the planning threshold is 42.105 orders. Real orders are integers, so at least 43 orders would beat the $230 baseline in this particular 1,000-session batch. Keep the unrounded rate for planning; rounding it to 4.2% slightly undershoots the target.

The 3.8% case increases revenue by 14% but reduces contribution after traffic by $46.80, or 20.35%. It is a useful warning against choosing the winning offer by revenue alone. These are arithmetic scenarios, not evidence that a coupon causes any particular conversion rate.

Check who pays and what can stack

Use the seller-funded amount from settlement records. A marketplace subsidy may lower the shopper's payment while leaving seller revenue unchanged; it cannot automatically be modeled like a seller-funded coupon. Likewise, free shipping can add a cost without reducing merchandise revenue. Keep those adjustments in distinct rows and inspect a test order before launching.

Shopify distinguishes product, order and shipping discounts, and combinations depend on eligibility and settings. Its official guidance also describes the calculation order. Verify your actual checkout because two visible promotion labels do not reliably describe the amount deducted from settlement. Record the final merchandise revenue, shipping revenue and all seller-funded reductions once each.

Explain the lift before repeating the offer

Extra orders can come from customers who would have purchased next week at full price. Compare the period after the offer ends, not just its busiest day. Segment new and returning buyers, examine bundle mix, and wait for refunds to mature. A promotion that attracts more low-contribution products may look stronger in total revenue while weakening the mix.

Separate a financial threshold from experimental evidence

The 4.2105% threshold answers what would have to happen. It does not establish that the offer produced a change. Thirty versus 38 orders is a small observation, and changes in audience, stock availability, creative or weekday can explain part of the difference. A fixed declaration such as “200 sessions prove the coupon works” would be unjustified.

Define the eligible audience and measurement window before testing. Where practical, randomize offer exposure and keep campaign treatment consistent. Record sessions, checkout orders, settled revenue, refunds and contribution for each group. Evaluate uncertainty alongside the financial threshold; a noisy apparent lift can still be too weak to justify a permanent lower price.

Choose an operating decision before extending the test

For this model, a measured 3.8% rate is financially below the required rate. The next decision could be a shallower discount, lower fulfillment cost or a clearer product page rather than a larger traffic budget. Write down the maximum test spend you can afford to lose, who can end the promotion, and when mature returns will be reviewed. Those limits belong to the business's cash position; they are not universal market benchmarks.