Fashion D2C Meta Ads: Inventory and Return Economics

For the surrounding account decisions, compare Subscription-Box D2C Meta Ads: Retention Economics Before Scale and use Fashion D2C Meta Ads: Inclusive Sizing and Fit Decisions as the next diagnostic.

In short

Rank fashion spend by the inventory that can still be fulfilled and the contribution retained after returns—not gross attributed revenue. The inputs are variant stock, fulfilled orders, matured returns, recognized revenue, variable costs, and campaign spend from your own systems.

The unit of control is product × variant × campaign over a declared cohort window. A style can look available while a core variant is exhausted; a campaign can look efficient before refunds and return handling reach the ledger.

Category economics as first-party inputs

• AOV = recognized revenue ÷ fulfilled orders.
• Variant depth = sellable units for a variant ÷ forecast eligible demand for the decision window. The forecast is internal and must expose its method.
• Return rate = returned orders ÷ delivered orders eligible for return measurement.
• Net units kept = delivered units − returned units for a matured cohort.
• Post-return contribution = recognized revenue − product cost − fulfilment − shipping − return handling − payment fees − discounts − allocated Meta spend.
• Contribution per campaign order = campaign post-return contribution ÷ fulfilled orders credited under the declared method. Attribution remains a reporting method, not incrementality.

Create a return reconciliation keyed to order and variant. Preserve order date, delivery date, eligibility date, return date, reason, refund, restock status, and variable handling cost. Do not apply one category return percentage to every product.

Creative and offer decision: advertise fulfilment-ready variants

Map each paid product set to live variant depth. Feature products whose pictured color and key sizes meet the internal stock floor. When a variant drops below that floor, change the eligible product set or creative according to the documented control; do not claim a platform action.

Use creative to answer product-specific uncertainty: fabric and construction evidence, dimensions, fit guidance, movement, pockets or closures, care, and the exact variant shown. Bundle or cross-sell hypotheses belong in separate cells with observed attach rate and contribution.

Measurement dictionary

• Paid ROAS = Meta-attributed revenue ÷ Meta spend.
• MER = total recognized revenue ÷ total paid-media spend.
• Return-adjusted campaign revenue = campaign-attributed recognized revenue − matured refunds under the same attribution method.
• Sell-through rate = units sold ÷ (units available at period start + documented receipts during the period).
• Stockout exposure rate = paid product-detail views for unavailable variants ÷ paid product-detail views with variant availability recorded.
• Post-return contribution rate = post-return contribution ÷ recognized revenue.

Illustrative operating model

Illustrative operating model — not a benchmark or expected result.

Input	Campaign A	Campaign B
Fulfilled orders	500	400
AOV	$80	$100
Recognized revenue	$40,000	$40,000
Product, fulfilment, shipping, fees, and matured return costs	$25,000	$28,000
Meta spend	$8,000	$8,000
Meta-attributed revenue	$20,000	$22,000
Sellable units in promoted variants	900	240

Revenue reconciles as 500 × $80 = $40,000 and 400 × $100 = $40,000. Paid ROAS is $20,000 ÷ $8,000 = 2.5× for A and $22,000 ÷ $8,000 = 2.75× for B. Post-return contribution is $40,000 − $25,000 − $8,000 = $7,000 for A and $40,000 − $28,000 − $8,000 = $4,000 for B.

The fictional example demonstrates why higher attributed ROAS does not settle the inventory-and-return decision. It is not a performance expectation.

Guardrails

• Reconcile feed and commerce availability at variant level and retain timestamps.
• Apply return costs only to cohorts mature under the same eligibility rule.
• Keep attributed revenue separate from total recognized revenue.
• Stop promoting a set when its internal stock-depth or contribution floor is missed.
• Treat forecasts and matched-period reads as estimates, not causal results.

Can software help?

Bach.ai audits your connected Meta account against 100+ checks, ranks what it finds by estimated impact, and proposes specific fixes. It stays read-only until you approve a change, then executes the approved change on Meta; connected Google Ads data is used for intelligence only. Think of it as an automated audit layer that surfaces issues and proposed fixes for your review — not a replacement for your team’s judgment, and it does not generate your creative.

Common mistakes

• Reading style-level availability while promoted variants are depleted.
• Comparing immature and mature return cohorts.
• Calling MER blended ROAS or using Meta spend as its denominator.
• Allocating budget from attributed revenue without variable costs.
• Treating a forecast sell-through value as observed demand.

FAQ

What should be the unit of fashion inventory reporting?

Use product × variant at the stock layer, then aggregate into documented paid product sets. Preserve the mapping so every campaign total can be traced back to sellable units.

When is a return cohort mature?

When every included delivered order has reached the same declared return-eligibility cutoff, plus the processing lag used by finance. Label earlier views provisional.

Should I optimize fashion campaigns to paid ROAS?

Use paid ROAS as an attribution diagnostic. Rank the economic decision with post-return contribution after product, fulfilment, shipping, return, fee, and acquisition costs.

How should bundles enter the model?

Measure (orders containing the add-on) ÷ (eligible base-product orders) as attach rate, then calculate contribution for the complete order. Test the offer rather than assuming an AOV lift.