Splitting MOQ Across Sizes and Colourways

Pet carrier production desk · Updated 2026-10-07 · 14 min read

Custom cat carrier shells come in soft, vacuum-formed and injection-moulded variants, and each one carries its own tooling and MOQ pool.

The minimum is measured in colourways, and that is not arbitrary

Brand teams usually hear "500 pieces" and assume it is an order minimum. It is not. It is a colourway minimum, and the distinction carries real money. Understanding why the number sits where it sits is the fastest way to plan a range that a production team can actually schedule.

Three costs reset when the colour changes. The first is the dye lot. A woven polyester or nylon body cloth is bought to a minimum run length, commonly somewhere between 300 and 600 metres depending on the mill and the construction, and a colour that does not reach that length is either bought at a premium or bought in excess and written off. The second is print setup, where a colourway carries its own screens, its own sublimation profile, or its own embroidery thread changeover. The third is line changeover: thread cones, webbing reels, binding tape and labelled components all have to be swapped and the first units off the line checked again.

None of those costs reset when the size changes. A large carrier uses more cloth than a small one, but it uses the same cloth, the same thread, the same webbing, the same hardware finish and the same packing specification. Cutting tables nest sizes together, so a size run is cut from the same lay. That asymmetry is the whole reason the minimum is written per colourway.

The implication for range planning is direct. If the commercial question is "how many variants can we afford", the answer is that sizes are nearly free and colours are not. A brand launching with a single 500-piece commitment can plausibly offer three or four sizes; offering three or four colours means three or four commitments, or a deliberate strategy to share a base material.

It is also worth being clear about what happens below the minimum. A 200-piece colourway is usually possible, and it is usually quoted at a meaningful premium per unit, because the fixed setup costs are spread over fewer pieces. Our production team will quote it honestly rather than decline it, but the brand should go in knowing that the premium is paying for a real constraint and not for a negotiating position.

Read the size ladder as an allocation problem, not a design problem

Once the colourway commitment is fixed, the next decision is how those 500 pieces divide across the ladder. This is a commercial decision disguised as a technical one, because the ratio determines whether the range sells out evenly or leaves a pile of one size.

Pet carrier demand is not uniform across a ladder. It clusters. For dog carriers the centre of gravity sits at the small-to-medium bands, which together typically account for the majority of units; the extra-small and extra-large tails are thin but must exist, because retailers treat an incomplete ladder as an incomplete range and because the tails are what get photographed. A defensible starting allocation for a four-size ladder is roughly 15 percent extra-small, 35 percent small, 35 percent medium and 15 percent large, and then adjusted by channel.

Channel matters more than breed data. A mass retail programme weighted to small-breed urban customers will push toward the small end; a specialist or online programme with a fit guide will sell more evenly; a programme sold through veterinary or grooming channels skews larger. The right input is not a national pet population statistic but the brand's own sell-through if it has any, or a comparable retailer's size mix if it does not.

Sizing should be expressed in weight bands rather than dimensions at this stage. A carrier specified as "up to 6 kg" is a buying decision the customer can make; a carrier specified as 46 x 28 x 28 cm is not. The dimensional spec follows from the weight band, and the two together give the pattern maker what is needed to grade the ladder. Where a programme covers both cats and dogs, run them as separate ladders rather than forcing one, because the aspect ratios genuinely differ.

Finally, do not allocate equally across sizes to keep it simple. Equal allocation is the single most common cause of a stranded size, and stranded sizes are worse than missing ones: they tie up cash, they occupy warehouse space, and they make the next reorder harder to plan because the brand orders around them instead of from them.

Four ways to spend one 500-piece commitment

Given a single colourway commitment, there are four defensible allocation patterns. They differ in how many SKUs they create, how much they cost per unit, and what they do to the next reorder. The table below holds the total at 500 pieces and shows what changes.

The last row is the pattern that looks attractive on a line sheet and costs the most to make. Five colourways at 100 pieces each is five dye-lot setups, five print setups and five changeovers, all amortised over a fifth of the volume. The unit cost index of 128 against 100 for the four-size single-colour ladder is not a negotiating artefact; it is the arithmetic of setup recovery.

The third row is where most of the value in range planning actually sits. Two colourways sharing a single neutral shell lot give the brand two colours on shelf, one dye lot to buy, and one fabric minimum to clear. It requires the colours to be introduced at a later stage — printing, panel substitution, or trim — which is a design constraint accepted at brief stage rather than a concession extracted later.

Where the material minimum binds harder than the piece minimum

The 500-piece figure is the visible minimum. Behind it sit three material minimums that bite in specific situations, and a buyer who knows them can plan around them.

Body cloth. Mills sell by length, not by garment count. A 600D polyester at roughly 1.5 metres of usable width will produce a certain number of panels per metre depending on the marker, and the mill's minimum order is expressed in metres. For a mid-size carrier the conversion works out somewhere in the region of 300 to 600 metres per colourway. Below that, the mill either charges a short-lot premium or the excess goes into dead stock. This is the constraint that makes a 100-piece colourway expensive.

Hardware. Buckles, sliders, D-rings and zipper pulls are bought from a hardware supplier in lots, and a custom finish — a specific anodised colour, a particular plating — carries its own minimum, often 1,000 to 3,000 pieces per finish per item. A brand that specifies four hardware finishes across a small range will hit this before it hits the fabric minimum. The fix is to standardise the finish across the range and let the colourway live in the cloth.

Print and labels. Woven labels, printed hang tags and rubber patches each carry a setup minimum, typically 500 to 1,000 pieces per artwork version. This is why a range with ten SKUs but one brand identity is cheaper than a range with four SKUs and four label artworks.

The general rule is that the minimum attaches to whatever is custom. Standard black webbing from stock has no meaningful minimum; webbing dyed to a brand Pantone reference does. Stock zippers have none; zippers with a branded pull do. Our production team flags these at quotation rather than at order placement, because the decision to use a stock component is a design decision and cannot be made once the sample is approved.

Material compliance interacts here too. A fabric certified to a restricted-substance standard such as OEKO-TEX STANDARD 100 is bought from a narrower set of mills, and dyed-to-order colours in certified cloth carry higher minimums than stock shades in the same construction. Abrasion performance, checked by a method such as ASTM D4966 for fabric abrasion resistance, narrows the field further.

Mixing rules: what can share a lot and what cannot

"Can we do 250 in one colour and 250 in another?" is the most asked question in range planning, and the honest answer is that it depends entirely on where the colour is introduced.

If the colour is in the body cloth, no. Two colours of cloth are two dye lots, and each clears its own mill minimum. There is no mechanism by which 250 pieces of one shade and 250 of another share a fabric purchase.

If the colour is in a printed panel, sometimes. A dye-sublimation print run has a setup cost but a much lower length minimum than a woven dye lot, so two printed colourways can in principle share a substrate. The practical limit is registration and waste: a print run for 250 panels carries a meaningful percentage of setup waste, and the unit cost of the panel rises accordingly.

If the colour is in the trim — webbing, binding, zipper tape, piping — yes, frequently. Trim is bought in shorter runs, and a colourway that is a neutral shell with a coloured webbing and a coloured binding is the cheapest possible way to create a second variant. This is the approach behind most successful two-colour launches.

If the colour is a lining change, also yes, subject to the lining mill's minimum, which is usually lower than the shell's because linings are lighter constructions sold in longer standard runs.

A useful way to frame it for a design team: the further downstream the colour enters the bill of materials, the cheaper the variant. Shell cloth is the most expensive place to differentiate; lining is cheaper; trim is cheaper still; print sits in the middle; and hardware finish is the most expensive of all because of the plating minimum. A range planned with that hierarchy costs less per variant and produces a cleaner production schedule.

What fragmentation actually costs, in numbers

Fragmentation costs money in four ways, and only one of them shows up on the quotation.

The first is setup recovery, already covered: fixed costs spread over fewer units. On a typical mid-complexity carrier, dropping a colourway from 500 pieces to 250 adds somewhere in the region of 6 to 10 percent to unit cost, and dropping to 100 pieces adds 20 to 30 percent.

The second is yield. Every changeover produces a run-in of units that are checked more heavily and occasionally rejected. On a line running seven-day schedules across 7 production lines, a changeover that costs forty minutes of output is trivial at 500 pieces and material at 100.

The third is inventory. Ten SKUs at 50 pieces each stock out unevenly and leave stranded sizes; four SKUs at 125 each sell through more predictably. The carrying cost of the stranded inventory over two seasons often exceeds the unit cost premium that the broader range was meant to avoid.

The fourth is the one that never appears in a quotation: attention. A programme with ten SKUs consumes roughly twice the quality-control, photography, listing and forecasting effort of one with four, and that effort is usually the scarcest resource in a young brand. Consolidating the range buys back planning capacity, which is worth more than the small unit-cost saving.

Set against that, there is a real benefit to breadth that should not be dismissed. Colour sells. A range with three colourways will convert better on a product page than the same range with one, and retail buyers ask about colour options early. The point is not to minimise colourways but to buy them in the cheapest possible way — shared shell, differentiated trim, one dye lot, staged launch.

Consolidation moves that protect the range

Four design moves reliably reduce the cost of a colour range without reducing its apparent breadth. All four have to be taken at brief stage; none can be retrofitted after a sample is approved.

Share the shell. Choose one neutral body cloth — charcoal, sand, or a mid grey — and let colour arrive elsewhere. Three colourways sharing one shell is one dye lot, one fabric minimum and one set of lab dips, against three of each if the shells differ.

Standardise the hardware finish. One finish across the whole range clears the plating minimum once. A brushed nickel or matte black works across colourways and reads as a deliberate family rather than a compromise. Reserve a second finish for a premium tier where the volume justifies it.

Grade the ladder, do not redesign it. A size ladder should be a graded version of one pattern, not four patterns. Grading changes panel dimensions and foam thickness; redesigning changes markers, hardware placement and packing dimensions. The first costs almost nothing to add sizes; the second costs a full sample cycle per size.

Stage the launch. Launch two colourways at 500 each, and hold a third for the reorder. The reorder benefits from an existing approved pattern and an existing dye-lot reference, so the third colourway arrives faster and cheaper than it would have as part of a fragmented first order. This sequencing is set out in more detail in our programme timeline, and the commercial side of per-unit cost under different allocations is worked through in the unit economics analysis.

Allocating across a three-season horizon instead of one order

The most useful change a brand can make to its MOQ planning is to stop treating each order as a standalone decision. A colourway minimum is a per-order constraint, but a range is a multi-season asset, and planning the two together removes most of the pain.

Consider a programme with a target range of three colourways across four sizes. Placed as one order, that is twelve SKUs and three colourway commitments, or 1,500 pieces minimum, plus the inventory risk of launching all twelve at once. Placed across three seasons, the first order is one colourway at 500 pieces across four sizes, the second adds a second colourway and refills the first, and the third adds the third colourway and refills both. The total commitment is the same, the peak inventory is roughly half, and the second and third colourways are ordered against real sell-through data rather than a forecast.

The mechanism that makes this work is that a repeat colourway does not need a full sample cycle. Once a pattern is approved and a dye-lot reference exists, adding a size or refilling a colourway is a scheduling exercise rather than a development one, and it can usually be slotted into the 35-50 day bulk window alongside other work. Adding a genuinely new colourway needs lab dips and a trim approval, which is weeks rather than days.

The risk to manage is the stranded size. If extra-small sells out in season one and extra-large does not, the season-two order should be planned around that, not around the original ratio. That means the allocation has to be revisited with actual sell-through before each reorder, and the SKU matrix revised rather than repeated. Brands that simply reissue the original matrix accumulate exactly the inventory they were trying to avoid.

A related point on seasonality: a limited-edition or seasonal colourway should be treated as a separate commitment with its own end-of-life plan, because a seasonal shade that sells through 60 percent leaves a colourway that cannot be refilled at minimum and cannot be mixed into a core order.

How the allocation should be written into the purchase order

An allocation agreed in email and never written down is the most common source of a disputed shipment. The purchase order needs to carry four things explicitly.

The SKU matrix. Every combination of size and colourway gets its own line with its own quantity. A four-size, two-colourway order is eight lines, not two. This is what the packing plan and the carton sequence are built from, and it is what makes a short shipment detectable.

The tolerance clause. Production overruns and underruns are normal. The standard position is plus or minus 5 percent per SKU, settled on actual shipped quantity, with the balance adjusted on the final invoice. Where a size must not fall below a threshold — a retailer minimum shelf quantity, for example — say so on the line, because a 5 percent underrun on a 75-piece size is four units and may matter.

The approval reference. Each line references the approved sample and its revision. A size that was graded after approval is a new revision and needs its own reference, otherwise bulk will be made to the superseded pattern.

The reorder position. State whether the order is a first production or a repeat, and if a repeat, reference the previous dye-lot number. That single line is what allows the next order to match colour rather than approximate it, and it matters more on the second order than on the first.

The inspection that releases the order follows the same matrix: sampling is per SKU within the lot, at AQL 2.5 general inspection level II. Where a colourway is split across two cartons, both are sampled. Our production team works to a documented allocation and revision process audited under ISO 9001, with BSCI social compliance at the SGS-verified production base. QUANZHOU JUNYUAN BAGS has run pet carrier programmes since 2014, its founder having been in the trade since 2004, across 7 production lines and 149 machines with monthly capacity of 200,000 units.