On-demand wins only when retained orders cover the complete cost

Production-route comparison needs a second model because sampling economics do not determine the cheapest way to make repeated orders. For route i, GFT defines total cost as C_i(q) = F_i + (v_i / y_i) * q, where q is accepted garments, F_i is fixed route or design setup cost, v_i is variable cost per started garment and y_i is accepted garments divided by starts. The starts required are q / y_i. This formulation makes recovery visible in the denominator instead of assuming that every start becomes saleable output.

The comparison must hold the artwork, substrate, size set, print placement, end-use quality and delivery requirement constant. Digital textile printing supports short runs and design flexibility, while rotary screen printing remains a relevant technical comparator with its own application knowledge and substrate requirements.[S01][S02] Neither statement supplies a universal cost ranking. A screen route may amortise design setup over a stable repeat, while a digital route may avoid some setup but carry different ink, pretreatment, curing and recovery costs. The valid evidence is a matched quotation and trial for the intended article.

Exhibit 37 shows an illustrative comparison. Screen setup is 900,000 XAF, variable cost is 4,200 XAF per start and accepted-output yield is 0.95. Digital setup is 180,000 XAF, variable cost is 5,300 XAF per start and yield is 0.90. These are invented model inputs. They do not describe any supplier or machine. Equal accepted volume is used on the horizontal sequence, so the different yields affect cost rather than the service delivered.

Setting the two illustrative equations equal gives q = (180,000 - 900,000) / ((4,200 / 0.95) - (5,300 / 0.90)), or about 490.5 accepted garments. Below that volume, the lower digital setup cost dominates in this case. Above it, the lower effective variable cost per accepted garment favours the screen route. The chart does not establish that either route passes the required physical tests. It shows only the arithmetic conditional on the assumed costs and yields.

A solver must test the denominator before reporting breakeven. If effective variable costs are equal and fixed costs differ, there is no finite crossover. If the calculated volume is negative, one route dominates throughout the non-negative production domain under the model. If the volume exceeds practical capacity or the demand horizon, the mathematical crossover is not an actionable result. A zero yield invalidates the equation and represents a route that cannot deliver accepted output.

Fixed cost also needs the correct repeating unit. Artwork setup may repeat per design, screen preparation per colour set, and machine setup per batch. A reorder using an unchanged approved specification should not automatically inherit the full first-design cost. Conversely, a material or artwork substitution may reopen setup and approval work. Separating new designs, exact repeats and revised repeats prevents a spurious route advantage created by inconsistent allocation.

An order can show a positive expected contribution and still run out of cash before collection. Deposits, blank or fabric purchases, consumable minimums, labour payment, packaging, dispatch, final settlement and refunds occur on different dates. Order-triggered production changes when work begins, but it does not remove those payment events or the stock held upstream.[S03] The cash model therefore belongs beside the contribution sheet, not inside a single margin percentage.

Exhibit 39 uses the same illustrative order as the worked example. The buyer pays a 180,000 XAF deposit on approval and 180,000 XAF on final collection. Material, ink and consumables are paid before production; labour, energy and setup follow; packaging, payment and delivery are paid at dispatch; and the assumed recovery cost occurs later. Cumulative cash is the sum of receipts less payments at each event. All dates and amounts are illustrative and do not establish customer terms.

The illustrative order ends with 60,000 XAF of cash surplus but reaches a cumulative deficit of 110,000 XAF before final collection. That deficit is the working-capital exposure in this simplified case. It can be larger when materials must be bought in supplier packs greater than the order requirement, when several orders share a production week or when settlement is delayed. The timeline should therefore be aggregated across all open orders rather than assessed one order at a time.

Commercial terms should make cancellation, approval, acceptance and remedy events explicit. Operationally, a refundable deposit is not equivalent to free cash while the related obligation remains open. The cash ledger should identify which receipts are restricted by the agreed terms and which expected payments remain exposed to refund or recovery. Responsible purchasing and production planning also matter beyond liquidity: OECD guidance identifies low wages, inefficient production planning and poor purchasing practices as common factors associated with excessive working hours.[S28] A cash-saving rush promise is not acceptable if it silently depends on unreasonable hours or unapproved subcontracting.

The final control is a reconciliation between three ledgers. The order ledger explains net retained sales and contribution. The capacity ledger explains what the complete route can deliver and what fixed period costs remain. The cash ledger explains when funding is needed. Capital structure, tax treatment and contractual obligations require separate qualified review; these examples are operational decision aids, not investment, legal or tax advice. Adoption should proceed only when all three views support the same bounded service under evidence that the pilot actually collected.