Salta al contenuto principale

MarginLab Case Investigations
Illustrative case
Case 03 · Full refund economics

What Happens When Refund Rate Rises from 5% to 12%

A seven-percentage-point refund increase reverses €5,600 more revenue. That is not the complete profit effect: some product value comes back, while acquisition, outbound delivery and payment fees have already been spent.

The business question

How much contribution is lost when 70 additional orders are refunded, after inventory recovery and return costs?

Initial evidence
CASE / 03
Refund rate5% → 12%50 → 120 orders
Extra refunded revenue€5,60070 × €80
Contribution decline€4,228€17,580 → €13,352
Cost per extra refund€60.40Under the stated recovery assumptions
The investigationCompare a refunded order with the retained order it replaces.
Investigation typeFull refund economics
Evidence basisConstructed dataset
Tax basisExcluding VAT/sales tax
OutcomeScenario-dependent
Methodology note: This is an illustrative ecommerce investigation, not a named merchant, observed customer result or guaranteed outcome. The cohort contains 1,000 one-item orders at €80. Refund rate means fully refunded orders divided by original orders. All refunds and product recovery are recognized within the comparison window.
Evidence 01 · Fixed cohort

Keep acquisition and shipment activity constant.

Both scenarios begin with the same 1,000 acquired and dispatched orders. Each sells for €80, costs €32 in inventory and incurs €7 outbound fulfillment, €2.40 nonrefundable payment fees and €18 acquisition spend.

Model assumptions
InputValueTreatment
Original orders1,000Same in both scenarios
Revenue per order€80Fully reversed on refund
Product cost€3280% of returns recover full cost value
Outbound fulfillment€7/orderNot recovered
Payment fees€2.40/orderAssumed nonrefundable
Acquisition spend€18/orderAlready incurred
Return shipping and handling€6/refundIncremental cost

The 80% recovery assumption means returned stock is saleable at its original €32 cost without impairment. The other 20% has no recoverable value. The €6 return cost includes processing and return transport; it is not added again as a separate restocking charge.

Evidence 02 · Order-level comparison

A refunded order can be loss-making despite recovered stock.

A retained order contributes €20.60. A refunded order has zero retained revenue but recovers an expected €25.60 of inventory value against its €32 original product cost.

Per-original-order contribution
ComponentKept orderRefunded order
Retained revenue€80.00€0.00
Original inventory cost(€32.00)(€32.00)
Inventory recovery credit€0.00€25.60
Outbound fulfillment(€7.00)(€7.00)
Payment fee(€2.40)(€2.40)
Acquisition spend(€18.00)(€18.00)
Return handling and transport€0.00(€6.00)
Contribution€20.60−€39.80
Difference caused by one additional refund€20.60 − (−€39.80) = €60.40

Equivalently: €80 reversed revenue + €6 return cost − €25.60 recovered inventory value.

Evidence 03 · Cohort reconciliation

The full contribution decline is €4,228.

The return rate rises by seven percentage points, from 50 to 120 refunded orders. The acquisition and original shipping totals do not change because the original cohort is unchanged.

Cohort economics; euros
Measure5% refunds12% refunds
Gross revenue80,00080,000
Refunded revenue(4,000)(9,600)
Net revenue76,00070,400
Product cost after recovery(30,720)(28,928)
Outbound fulfillment(7,000)(7,000)
Payment fees(2,400)(2,400)
Acquisition(18,000)(18,000)
Return costs(300)(720)
Contribution17,58013,352

The bridge is €5,600 more reversed revenue plus €420 extra return costs, less €1,792 additional inventory recovery: a €4,228 decline. Contribution falls approximately 24.1%, more sharply than the revenue reversal alone might suggest.

Evidence 04 · Recovery and cash

Recovered inventory is an asset, not an immediate cash refund.

The accounting model credits saleable stock when it returns. Cash is recovered only when that stock is sold again or otherwise monetized. A slow resale cycle can therefore create more liquidity pressure than the contribution bridge shows.

Finding 1

Cash already spent

The merchant cannot undo the original €18,000 acquisition or €7,000 outbound fulfillment simply because an order is refunded.

Finding 2

Product value at risk

If returned items cannot recover their cost, the inventory credit is too optimistic. Inspect condition, repackaging and expected resale proceeds.

Finding 3

Reporting timing

Compare mature cohorts. Current-month refunds may relate to prior-month orders and distort the apparent relation between acquisition and returns.

IAS 2 addresses inventory cost and recoverable value; the standard summary supports the distinction between stock recognition and the later cash event. The numerical recovery assumptions here are hypothetical.

Decision test · Recovery sensitivity

The value of prevention depends on the recovered product.

For the 70 additional refunds, changing the recovery rate changes the damage. Hold all other assumptions constant to isolate this uncertainty.

Sensitivity for the 70 extra refunds only
Full-cost recovery rateRecovered value/refundLoss per extra refundTotal decline
100%€32.00€54.00€3,780
80% base case€25.60€60.40€4,228
50%€16.00€70.00€4,900
0%€0.00€86.00€6,020

At the base recovery rate, preventing 20 refunds would preserve €1,208 contribution before the cost of the intervention. That provides a budget ceiling for a narrowly defined test, not permission to obstruct legitimate returns.

Validation · Address the source of the refunds

Investigate causes before changing the policy.

The useful intervention improves the purchase or product experience. A return-friction change can move the reported rate while damaging conversion, repeat demand or customer trust.

  1. Segment the additional 70 refunds

    Compare product, acquisition promise, size or fit, damage, delivery and customer cohort. Find the concentration before choosing a remedy.

  2. Test a cause-specific intervention

    For example, improve sizing information or packaging for the affected SKU. Track the cost and mature refund change in comparable groups.

  3. Measure the complete result

    Include conversion, contribution, return processing and customer support. Confirm that avoided refunds represent better outcomes rather than delayed reporting.

Illustrative prevention economics20 avoided refunds × €60.40 − €500 intervention cost = €708

This is a conditional improvement. The test must demonstrate the 20-refund reduction without offsetting damage elsewhere.

Use the contribution margin framework to avoid treating refunds as a revenue-only adjustment.

Case conclusion · quantified decision

Seven extra refund points remove €4,228 of contribution.

The full result combines lost sales, additional return costs and recovered inventory. Acquisition, delivery and payment costs remain embedded in the refunded orders rather than disappearing with the revenue.

Unit finding

Each extra refund costs €60.40.

It replaces a €20.60 retained-order contribution with a €39.80 loss under the stated assumptions.

Decision impact

Twenty prevented refunds can fund a focused test.

A €500 intervention that genuinely avoids 20 refunds improves contribution by €708 before any other changes.

Use mature cohorts and a realistic inventory-recovery assumption, then budget prevention against the contribution it can preserve.
Continue the investigation

Recalculate profit per original order

Include refunded orders in the economic boundary so customer acquisition and shipment costs remain visible.

All calculations are illustrative. The decision depends on the stated cost, demand and timing assumptions.