Take your existing per unit landed cost, product, freight in, insurance, duty, brokerage, receiving, divided by units, and add one term: expected return cost per unit sold. Build it from three numbers your operation already knows or can measure in a month: the category's returns rate, the average cost of the return journey including the label, the carrier, and the receiving inspection, and the recovery rate, the share of returned units that go back to sellable stock at full value. The unsellable share carries its full landed cost into the loss column, and pretending otherwise is how margins surprise people in January.
Allocate it per unit sold rather than per unit returned. Pricing has to recover the return leg from the customers who keep the product, because the customers who return it, in a free returns model, are not paying for the trip. That allocation is what turns returns from a customer service metric into a pricing input, which is where it has belonged all along.
Duty paid at import is not automatically refunded when the goods leave again. In the US, duty drawback under 19 U.S.C. 1313 can recover up to 99 percent of duties on qualifying exported or destroyed goods, including in defined unused merchandise and rejected merchandise situations, with claims, records, and timelines attached. Other markets run their own reliefs for re exported returns, and some flows, particularly low value consumer returns, cost more to document than the duty at stake.
The model's job is to carry the realistic recovery, not the theoretical one. If your returns operation consolidates and re exports with drawback claims filed properly, price the recovery in, minus the broker and administration cost of claiming. If returns are destroyed locally or liquidated in market because the freight home exceeds the value, the duty is a sunk cost of that unit and the model should say so plainly.
Sellers budget the return leg at the outbound parcel rate and discover the differences in the invoices: consumer return labels at retail or semi retail rates rather than the negotiated outbound contract, dimensional weight on goods repacked worse than the factory packed them, residential pickup and surcharge profiles the outbound flow never met, and cross border returns freight that prices like an import because it is one. Measure the leg from actual return invoices for a quarter before trusting any per unit figure in a pricing deck.
Category differences swamp store averages here. A category returning at five percent in compact packaging and a category returning at thirty percent in bulky packaging are not two versions of one problem. Build the return cost term per category at minimum, per product family where the data supports it, and revisit it when packaging or sizing content changes move the returns rate, because the term should move with them.
Once the return leg has a number, decisions that looked like service choices price themselves. A sizing tool, a better size chart, a packaging change that cuts damage returns, each can be valued in reduced return cost per unit sold and compared honestly against its own cost. Stores that run this find the cheapest returns programme is the one that prevents the return, and they can finally prove it in the same units as every other claim on the budget.
Charge the term visibly in the model, reviewed quarterly alongside freight rates, rather than burying it in a general margin haircut. Buried assumptions do not get maintained. A named line with an owner gets recalculated when rates and returns move, and in a business where both move every season, maintenance is most of the accuracy there is.
Multiply the category returns rate by the full cost of a returned unit's journey and handling, including the unsellable share's lost landed cost, and allocate that across all units sold. Measure each input from your own invoices and returns data for a quarter.
Sometimes. US drawback can recover up to 99 percent of duties in qualifying situations, and other markets have their own reliefs. Whether it pays depends on values, documentation, and volumes, which is a broker conversation worth having with your returns data in hand.
Free returns change who pays, not what it costs. The return leg still happens and still needs funding from the units that are kept, so the cost term belongs in the model regardless of the customer facing policy.
Quarterly, or whenever freight contracts, packaging, or a category's returns rate moves materially. The term is an average of moving parts and goes stale on the same schedule they do.