Kidswear operations: size curves, sibling packs, and inventory truth across channels

Why does kidswear inventory break the size curve model that works for adult apparel?

A kidswear ops lead I worked with last quarter opened her Monday with two problems on the same style. The 3T in the hero legging was oversold by 40 units across Shopify and a Faire order that landed Friday night. The 6M in the same style had 220 units sitting in the 3PL, untouched for six weeks, and her buyer had already flagged it for markdown. Same SKU family, same warehouse, same week. One size was starving the channel, another was aging into a write-down. Her planner had built the buy against a size curve that assumed the demand shape she saw last spring. The shape had moved.

This is the kidswear inventory size curve operations problem in one scene, and it is not a planning failure in isolation. It is a BP3 problem, inventory truth getting weaker at exactly the moment when the number of SKUs per style, the channel mix, and the pack logic all start compounding.

What does kidswear inventory size curve operations actually mean?

Kidswear inventory size curve operations is the discipline of buying, allocating, replenishing, and reporting on inventory where a single style carries far more size variants than adult apparel (often 12 to 18 across newborn, infant, toddler, and kids age bands), where those sizes do not sell in a stable ratio, and where the same pool of units has to service DTC singles, wholesale prepacks, sibling bundles, and sometimes a marketplace feed on top. The operational unit is not the style. It is the style-size-channel triple, and the truth of that triple has to hold across the ERP, the 3PL, the Shopify storefront, and every wholesale portal at the same time.

That definition matters because most inventory tooling was built for the style-size pair. Kidswear needs the third dimension and does not get it, so the reconciliation work moves into a spreadsheet and a person.

Why does the size curve drift faster in kidswear?

From the vendor evaluations I sit through with kidswear brands in the $10M to $30M band, the same pattern shows up in the demand history almost every time. The curve the planner used to place the PO looks nothing like the curve the season actually sold. A few structural reasons.

First, grow-out compresses the sell-through window per size. A 12M does not have a twelve-month customer. It has a roughly three-month customer, and if the drop arrives late or the size mix is off, the size ages out before it clears. Adult apparel forgives a slow M. Kidswear does not forgive a slow 18M.

Second, the wholesale channel places prepacks that assume a curve, and the DTC channel sells singles against a different curve. A prepack of 2-2-2-1-1 in 2T-3T-4T-5T-6T pulls units in a shape the DTC customer did not ask for. If ATS is calculated globally, DTC starts selling sizes that were quietly earmarked for the next wholesale ship window, and the chargeback shows up eight weeks later.

Third, sibling bundles and gift sets pull matched sizes from separate style pools. A two-pack that promises a 2T top and a 4T bottom for older-sibling gifting reserves inventory in a shape no single-SKU report will surface. When the buyer looks at 2T stock on Monday, the number is wrong by the count of open sibling-pack orders that have not yet picked.

How does the channel mix make the truth worse?

A kidswear brand at $15M running wholesale, DTC, and a 3PL is doing the reconciliation work we have described elsewhere as 6 to 9 hours per week of someone comparing Shopify inventory, the 3PL’s on-hand file, and the open wholesale order book. In kidswear, that number understates the pain because the reconciliation happens at the size level, not the style level. If a style has 14 sizes and the brand carries 300 styles, the ops team is not reconciling 300 rows. They are reconciling 4,200 rows, and the mismatches cluster in the sizes that matter most (the middle of the curve, where wholesale prepacks and DTC demand collide).

The oversell rate at peak that we see in the $15M cohort, roughly 2 to 3 percent, is not evenly distributed in kidswear. It concentrates in three or four sizes per hero style. Those are also the sizes with the highest gross margin recovery on a full-price sale, so the oversells cost more per unit than the top-line number suggests. The remedy people reach for first (a static safety stock buffer per SKU) makes the problem worse, because it locks units away from the channel that is actually converting them.

The structural answer is a channel-aware available-to-sell calculation, computed against committed wholesale orders, open prepack components, sibling-bundle reservations, and a per-size safety floor that is set by velocity, not by a flat percentage. That is what connected inventory across DTC, wholesale, and 3PL is supposed to do at BP3. When it does not, the ops team does it by hand, and the hand is slow.

Where do sibling packs and prepacks break the standard inventory model?

Sibling packs are the kidswear-specific version of a kitted SKU, and most mid-market inventory systems treat kits as an afterthought. The pack has its own SKU for the storefront and the wholesale linesheet, but the fulfillment is against component SKUs that live in the same pool as the singles. Three things go wrong.

One, the pack is oversold because ATS on the pack SKU was calculated from a stale snapshot of the components. The customer sees stock, buys the pack, and the pick ticket fails at the 3PL because one component is already committed.

Two, the pack is undersold because the system will not commit the pack unless all components are in the same location, and the components are split across two warehouses. The pack shows out of stock on the site even though the total inventory exists.

Three, the pack is fulfilled but the accounting does not know how to allocate the revenue and the cost across the components, so gross margin by style is wrong. Finance closes the month with a number the merchandiser cannot reconcile back to the sell-through report.

Wholesale prepacks add a fourth failure mode. A prepack sold on a linesheet in July for a September ship reserves component sizes for two months. If those reservations are not visible to the DTC channel, DTC oversells against inventory it does not actually have. If those reservations are visible but held at the style level rather than the size level, DTC undersells the sizes that are not in the prepack. Both are common. Both come from the same root cause, which is that the ATS calculation is not aware of the pack decomposition.

When does the buying decision go wrong?

The buying decision for a kidswear season starts with a size curve, either historical or aspirational, and multiplies it against a total unit buy per style. In practice, the historical curve is a blend of channel curves that were never separated, so the buyer is optimizing against a shape that does not exist in any single channel. The DTC curve skews younger (more newborn and infant, because that is where the gifting sits). The wholesale curve skews older, because the buyer at the specialty retailer is filling out a floor set that a parent will shop with a walking child. Averaging the two produces a curve that fits neither.

The point of view I would push here is that kidswear brands should build and buy against separate channel curves for hero styles, then reconcile them against the manufacturing minimums, rather than blending first and buying second. The size curve is not a property of the style. It is a property of the style-channel pair, and treating it otherwise is the reason the 3T oversells and the 6M sits.

Running open-to-buy weekly during selling season, not monthly, is the other half of the same discipline. Kidswear seasons are short. A monthly OTB cadence in a category where a size can age out in twelve weeks is a cadence that finds the problem after the markdown.

How should returns feed back into the size truth?

Kidswear returns run higher than most adult categories, partly because parents order two sizes to find the fit, partly because gifting produces a wave of sizing exchanges in the first two weeks of a life event. If those returns take three weeks to post back to inventory as available (which is common when the 3PL grades, the QC bench inspects, and the ERP updates on a batch), the size truth is wrong for the entire window. The buyer sees 3T as sold through and reorders. The units come back into stock the week the reorder PO lands. Now the brand is long on 3T and the markdown clock starts.

The operational rule that fits kidswear here is that returns should post to inventory in days, not weeks, and the grading logic should be simple enough to run on the same shift the return arrives. Complex disposition rules that require a merchandiser sign-off are the reason the posting is slow. The fix is a default disposition per SKU class with an exception queue, not a case-by-case review.

What does BP3 look like specifically for kidswear?

The inventory truth scorecard that anchors BP3 assessment is a useful frame for a kidswear ops team because it forces the question at the right level of granularity. In an adult apparel brand, the scorecard asks whether the on-hand number in the ERP matches the 3PL and matches what the storefront is selling against. In a kidswear brand, the same questions have to be asked per size, per channel, and per pack decomposition.

A passing BP3 grade in kidswear means the ATS shown to Shopify at 9 AM Monday is calculated against committed wholesale ship windows for that week, open prepack components, open sibling-pack orders, and a per-size safety floor that reflects the last four weeks of velocity, not a flat percentage. It means the pack SKU is not oversold because the component snapshot is stale. It means returns land in the available pool within 48 hours of arrival at the 3PL. Very few mid-market kidswear brands can answer yes to all four without a manual step, and the manual step is the FTE we keep describing as data plumbing.

Why do vendor evaluations miss this?

The comparison conversations kidswear brands walk into during platform selection almost never surface the pack decomposition question or the channel-aware ATS question, because the demos are built around adult apparel workflows. The buyer sees a clean size matrix, a wholesale portal, a Shopify sync, and assumes the pieces will hold at their SKU count. They will not, unless the ATS engine understands packs and channel commitments as first-class inputs.

The evaluation question that separates platforms cleanly is this. Show me the ATS for a hero style with 14 sizes, three of which are inside an active wholesale prepack shipping in two weeks, and two of which are components in a sibling-pack SKU that has 30 open DTC orders. Show me what Shopify sees, what the wholesale portal sees, and what the 3PL sees, all at the same timestamp. Most platforms cannot show that view without a spreadsheet in between, and that is the tell. The Shopify wholesale inventory sync question is not a technical integration question. It is a data model question, and kidswear is where the model gets stress-tested first.

The size that oversells is the size that funds the season

If a kidswear ops team takes one thing from a BP3 diagnostic, it should be that the oversells and the aged inventory are the same problem seen from two sides. Both come from a size curve that was built once and never refreshed against the actual channel demand, and from an ATS calculation that treats a global on-hand number as if it were the answer to the question the channel is asking. It is not. The channel is asking a per-size, per-channel, per-pack question, and the answer that fits adult apparel does not fit kidswear.

The brands that get this right treat the size curve as a living artifact per channel, run OTB weekly during the compressed selling window, close returns to inventory in under 48 hours, and pick a platform whose ATS engine can decompose packs without a spreadsheet in the middle. The brands that do not get this right pay for it in the two sizes per hero style that finance did not forecast, and in the markdown line that shows up at the end of the season looking larger than the plan said it would.

FAQs