A percentage discount gives away margin on every discounted unit. A bundle can sometimes increase order value while spreading fixed fulfilment and payment costs across more units. The right comparison is contribution per order and contribution per visitor—not headline revenue.
promotion contribution = order revenue − product cost − fulfilment − shipping subsidy − fees − expected returns − acquisitionCompare the promotion mechanics
A 20% discount lowers the revenue of an order that might have happened anyway. A bundle may require extra product cost but can raise AOV and reduce the effective burden of fixed per-order charges. Both can fail if the incremental conversion does not offset the contribution surrendered.
Worked example: 20% off vs two-unit bundle
Suppose one unit normally sells for $50, carries $18 of direct product cost and $9 of other variable order costs. Baseline contribution is $23. A 20% discount reduces revenue to $40 and contribution to $13. A two-unit bundle at $85 with $36 product cost and $11 of other variable costs leaves $38 contribution.
| Offer | Revenue | Modeled contribution |
|---|---|---|
| Single at full price | $50 | $23 |
| Single at 20% off | $40 | $13 |
| Two-unit bundle at $85 | $85 | $38 |
Translate contribution loss into required sales lift
If contribution falls from $23 to $13, you need roughly 77% more discounted orders to produce the same total contribution as the baseline, before considering any change in acquisition cost. That is a much harder hurdle than simply “selling more units.”
Decision rule
Prefer the promotion that produces more expected contribution per visitor at an acceptable inventory and return risk. Model each option in the Discount Profitability Calculator, then test one offer at a time.
Use your own numbers
Every worked number on this page is illustrative. Replace it with current data from your own statements, invoices, supplier quotes and operating history. Profit Per Order calculators keep assumptions editable so a content example never becomes a hidden business rule.
For a paid decision layer, inspect the public Profit Action Report sample before considering the one-time $19 report.
Compare offers on the same traffic base
Assume 1,000 comparable visitors and a 2% baseline conversion rate. Twenty full-price orders at $23 contribution produce $460. The $13 discounted contribution would require 35.3846 orders, or 3.5385% conversion, to match it in expectation. Since actual orders are whole, 36 orders are needed to exceed $460. This is the source of the 76.9% relative conversion lift; it is not a prediction of customer response.
At $38 contribution, the bundle needs 12.1053 orders per 1,000 visitors, or 1.2105% expected conversion, to match the same $460. At a tested 1.4% conversion, 14 bundles would produce $532 and ship 28 units. That is more contribution but also more inventory consumed than the 20-unit baseline. The CSV keeps orders and units separate so the bundle is not counted as two customers.
Where the simplified comparison can fail
The original example assumes fixed totals for the other variable-cost line. Real percentage payment fees will change with revenue, a larger parcel can cross a carrier tier, and picking the second unit may incur an extra charge. Recalculate those lines instead of assuming the $11 bundle allowance will hold for every basket.
Track conversion, contribution per visitor, items shipped, returns, stock availability and shipping cost across comparable cohorts. Keep traffic quality and promotional timing comparable. The test needs enough observations to distinguish a repeatable effect from noise; the 1,000 visitors here are arithmetic scaffolding, not a universal sample-size requirement. Do not declare a winner before checking later refunds.