A seasonal buy needs a recovery plan before the first unit is ordered

Delaying a decision is not the same as delaying physical differentiation. Postponed manufacturing keeps material in a more generic form until a better demand signal arrives, then converts it into a specified product. Postponed planning delays the commercial or production decision itself. The first can preserve feasible options. The second can simply move commitment closer to the deadline, increase urgency or transfer stock risk to another party.

The Japanese apparel study makes this distinction empirically important. In its survey of Japanese apparel wholesalers, postponed manufacturing was associated with lower inventory, while postponed planning increased inventory when firms externalised the retail function. The authors also identify retail controllability and the accuracy of additional-production decisions as conditions to consider.[S17] These are conditional associations from that sample, not a causal forecast for a Cameroon factory or a promise that every postponement programme reduces stock.

For an apparel maker, candidate manufacturing postponements include holding an approved common fabric, delaying a print or keeping a label generic until the order specification is confirmed. Each candidate requires a technical test. The base material must genuinely serve the later variants; conversion equipment and skills must be available; setup and changeover must fit the remaining window; and final inspection must still leave time for the promised delivery. Material planning and item-version controls provide the specification discipline, but they do not establish that a particular late-stage process is feasible.[S07]

The economic comparison should follow the decoupling point. Generic stock may serve several later demands, but conversion adds cost and can fail. Finished stock avoids that later conversion risk while carrying more variant-specific residual exposure. Postponement is worthwhile only when the expected reduction in mismatch and terminal loss exceeds the additional conversion, reservation, coordination and service costs under common demand scenarios. No universal inventory reduction should be inserted into that comparison.

A delayed planning decision needs an equally explicit owner and information rule. If a retailer, distributor or customer controls the signal used for additional production, the factory must know whether that signal is a forecast, a reservation or a binding order. Otherwise planning postponement can produce repeated speculative commitments while each party believes another owns the residual. Contract definitions and call-off liabilities should be settled before the season, not during an urgent second buy.[S08][S17]

GFT would therefore approve postponement as a timed operating design rather than a general principle. The record would name the generic stock object, permissible later variants, latest conversion start, usable-completion deadline, responsible approval and fallback route. WP20 owns the broader process and product-design choices. This chapter retains the narrower inventory test: the option must remain technically convertible and commercially useful after the demand information arrives.[S07][S17]

Residual recovery is a process with time, cost and uncertain eligibility, not a single salvage rate. A markdown can accelerate sales but changes revenue and can alter demand for other variants. A customer return does not immediately become usable stock. It must be received, inspected, graded and, where necessary, repaired, cleaned or repacked before it can re-enter the selling process. European Environment Agency evidence describes complex return handling and potentially significant effects on resale timing and value.[S26] Its European return rates and destruction estimates must not be transferred to Cameroon.

The recovery calculation should follow each route separately. Expected net resale recovery is the expected selling proceeds less incremental discount support, handling, repair, repacking, channel and disposal costs, with the timing and probability of each outcome stated. An item that misses the remaining selling window may have a different route from an item returned early. A damaged return may have no resale route even though an undamaged return can be restocked. Combining these conditions into one average salvage price hides the decisions that create the value.

Returns also require a stock-state control. Until inspection establishes condition and ownership, the returned unit should not inflate promiseable stock. If it becomes eligible for resale, its release time belongs in the remaining-window calculation. If it is repaired, the repair capacity and additional cost belong in the same comparison. A return may reverse revenue commercially while still failing to restore a unit that can be sold again in the season.

Accounting and decision economics remain distinct. IAS 2 describes inventories as measured at the lower of cost and net realisable value, with net realisable value reflecting estimated ordinary selling price less completion and necessary selling costs.[S09] That measurement does not prove that a buyer exists or that the estimated proceeds will be realised. The seasonal order model needs expected incremental outcomes, while the financial record needs the applicable accounting assessment. Local professional review is required; this chapter is not accounting or legal advice.

An EU-facing scenario adds a scope test. The European Commission states that, from 19 July 2026, large companies across the EU are prohibited from destroying unsold clothes, clothing accessories and footwear, subject to specified circumstances and exemptions.[S21] This is not a blanket obligation on a Cameroon manufacturer. The entity, company size, market, product and contractual role must be established by legal review, with WP17 providing the detailed interpretation.

The Commission's explanation prioritises selling through discounts or alternative markets, donation, and preparation for reuse through repair, refurbishment or remanufacturing, while restricting destruction to specified cases.[S21] Exhibit 30 converts that principle into an illustrative factory decision flow. The flow is a proposed control design, not a claim that every route is locally available or legally sufficient.

The stop rule should be written before the initial order, while the cost of another unit can still be assessed dispassionately. It should name the last order date, last useful receipt, demand-evidence refresh, residual exposure limit and role authorised to approve an exception. Without those commitments, every new sales narrative can justify another buy after the feasible selling window has narrowed. The single-period model supplies the marginal economics, while inventory measurement and recovery records supply the terminal exposure.[S06][S09]

At each review, the planner should project terminal stock under declared demand and usable-receipt scenarios. The exposure includes current usable stock, firm inbound supply and any proposed order, less sales expected before the window closes. Variant-level residuals must remain visible. The review then compares expected net recovery, handling cost and cash committed with the approved limit. A statistically attractive reorder that breaches the limit requires an explicit business decision rather than an automatic exception.

The action ladder begins with preventing additional mismatch. Stop or reduce uncommitted orders where the marginal case fails, convert only where specifications and capacity permit, and do not count an unreliable cancellation right as certain relief. Commercial actions can then target price, channel, repair, resale, donation or reuse according to actual eligibility. Final disposal requires documented ownership and lawful route review. The appropriate sequence depends on time and net recovery, not on a universal preference for markdowns.[S21][S26]

A stop rule must also avoid the opposite error. Ending all replenishment on a fixed calendar date can sacrifice profitable demand when a fast, qualified option still clears its economic and cash tests. The rule should therefore stop unsupported exposure, not every later order. A second buy may proceed when fresh requested-demand evidence, usable-receipt timing, variant fit and residual exposure all remain within the approved conditions. Governance should make both continuation and cessation explainable.

The final evaluation should compare the original assumptions with realised outcomes without rewriting history. Which demand scenarios were available at approval, which receipts were expected to be usable, and which recovery values were authorised? Actual contribution, terminal stock, cash recovery and service should be reported against that record. A high sell-through achieved by early stockouts is not unqualified success, and a low residual achieved through indiscriminate discounting may conceal lost margin. The policy earns renewal when its complete seasonal result is understood.