Apparel Landed Cost per SKU from Pre-Pack Assortments

An apparel pre-pack assortment turns one supplier invoice line into several inventory-costing rows. A carton described as 1S/2M/2L/1XL for one style and color is not one SKU in the receiver. It is six units spread across four size-level SKUs, each of which needs its own unit cost before inventory posts.

That is why apparel landed cost per SKU pre-pack assortment work has three steps that generic landed-cost guides usually skip: expand each pre-pack line into one row per style, color, and size; allocate freight and shared charges by a defensible basis, usually FOB value or volume for apparel imports; then allocate duty from the customs entry line, not from a shipment-wide average rate.

Take a common receiving-day example. A 40-foot ocean container arrives with 12 styles, 4 colors, and 5 sizes, so the inventory system needs 240 SKU rows. The supplier commercial invoice shows only 48 lines because each line is style by color with size quantities across columns. The ocean carrier invoice adds $4,200 of freight. The customs broker invoice adds $850 of brokerage and processing charges. CBP Form 7501 shows $9,400 of duty across MFN duty, Section 301, and temporary Section 122 layers.

The inventory accountant's job is not just to calculate a shipment total. The job is to put the right landed unit cost onto every style-color-size row so the ERP receiver, inventory valuation, and season margin report all agree. A blended duty rate or a freight allocation done before the pre-pack is expanded will make some sizes or styles look more profitable than they are and push the error into every sale that follows.

Classify the packing unit before touching the math

The first branch is vocabulary, not arithmetic. Apparel teams use case pack, pre-pack, and bundle to describe three different packing patterns, and each one changes the landed-cost workpaper.

A case pack is multiple units of one SKU in the same carton: for example, 12 units of style 2841, black, size M. The invoice line is already at the SKU grain or close to it. You still need to allocate freight, duty, brokerage, and other import costs, but you do not need to explode the line across sizes first.

A pre-pack is a fixed size run for one style and color: for example, 1 small, 2 medium, 2 large, and 1 extra large. The supplier invoice may show one line for the pack, but the inventory system receives four SKU rows. The costing file has to split the pack into size-level rows before any landed-cost component can be attached correctly.

A bundle is a mixed-style assortment. It is less common in wholesale receiving than case packs and pre-packs, but it is more dangerous when it appears because one supplier line can contain multiple styles, colors, or product families. A bundle usually needs an assortment breakdown before cost can be assigned. If that breakdown lives only in the PO, packing list, or merchandising file, the landed-cost worksheet has to pull it in before allocation.

This is the apparel-specific fork that separates the workflow from the broader wholesale distributor landed cost workflow. A distributor line for 100 identical replacement parts can move straight to allocation. A pre-pack apparel line cannot, because the unit that was invoiced is not the unit that will be stocked, sold, or margin-tested.

Normalize the supplier invoice into one row per style, color, and size

Most apparel pre-pack landed-cost mistakes start in the supplier invoice layout. Overseas manufacturers often invoice in a matrix-across-columns format: one row for a style and color, columns for S, M, L, XL, and XXL quantities, one FOB unit cost, and one extended line total. That layout is efficient for the factory, but it is not the row structure an apparel ERP needs for receiving cost.

The normalization step is an unpivot. For each supplier invoice line, every nonzero size column becomes a row in the working file. A line for style D142, color ivory, with 20 small, 40 medium, 40 large, and 20 extra large units becomes four rows, each carrying the same style, color, supplier invoice number, PO, FOB unit cost, origin, and line reference. The size and quantity change by row. Extended FOB is recalculated as quantity times FOB unit cost.

The target file should look like the receiver's grain:

• Style
• Color
• Size
• SKU or ERP item code
• Quantity
• FOB unit cost
• Extended FOB value
• Supplier invoice number and line reference
• Carton, pack, or shipment reference where available
• Country of origin and HTSUS reference when already known

Some supplier invoices are cleaner: one row per size variant. Those do not need matrix expansion, but they still need validation. Quantities should tie back to carton counts and invoice totals, style and color codes should not be normalized into new labels, and zero-quantity sizes should not create SKU rows just because the matrix column exists.

Invoice Data Extraction fits before the allocation math. The product can turn supplier invoice PDFs or images into structured Excel, CSV, or JSON output, which is useful when the invoice is a matrix layout, a size-variant layout, or a mixed batch of both. The extraction prompt should ask for the same fields fields a commercial invoice data extractor should capture for costing: style, color, size columns, quantities, FOB unit cost, extended amount, invoice number, PO, currency, origin, and any line reference that will later match to the customs entry.

The control is simple: the expanded rows must sum back to the supplier invoice. If the original line says 120 units at $8.50 FOB, the exploded size rows must still total 120 units and $1,020 FOB. If they do not, do not allocate freight or duty yet. Fix the expansion first, because every downstream cost uses that row set.

Build the landed-cost stack with apparel's 2026 duty reality

The landed-cost stack for an apparel import usually starts with FOB product cost, then adds international freight, duty, customs brokerage, Merchandise Processing Fee, Harbor Maintenance Fee for ocean shipments, terminal handling, inland delivery, and insurance. The accounting question is which of those costs should be capitalized into inventory and how each one should be assigned to the SKU rows produced by the expansion step.

As of May 6, 2026, the customs-fee pieces are worth treating explicitly. Under 19 CFR 24.23 Merchandise Processing Fee regulations, merchandise that is formally entered or released is subject to an ad valorem Merchandise Processing Fee of 0.3464 percent of declared value. CBP's FY 2026 customs user fee adjustment keeps that ad valorem rate and sets the formal-entry minimum at $33.58 and maximum at $651.50, effective October 1, 2025. For ocean cargo, the Harbor Maintenance Fee is separate; 19 CFR 24.24 sets it at 0.125 percent of the value of commercial cargo loaded or unloaded at covered U.S. ports.

Apparel duty is usually the largest source of distortion. Many garments in HTSUS chapters 61 and 62 carry MFN duty rates in a much higher band than importers see for industrial supplies. A women's synthetic woven dress under heading 6204.43, for instance, can sit at a 16 percent MFN rate depending on the exact 10-digit classification in the USITC Harmonized Tariff Schedule. Other woven and knit garment lines can land in the 10 percent to 32 percent range before trade-remedy layers.

That is why the cost stack should not be collapsed into one shipment percentage. Ocean freight may be allocated by value or CBM. Brokerage may be spread by entry or by line count. MPF may be capped at the entry level and then allocated. HMF applies only to vessel cargo. Duty has to follow HTSUS line treatment. When freight forwarder invoice processing for ocean and air shipments produces one total and the customs entry produces several duty lines, the landed-cost file has to preserve those differences rather than average them away.

Choose an allocation basis that matches the charge

After the pre-pack is expanded, the next decision is allocation basis. The right basis depends on the charge, not on whatever column is easiest to use in the spreadsheet.

By FOB value is often the default for apparel shipments with wide first-cost variation. If a container mixes silk dresses, embellished jackets, cotton T-shirts, and basic leggings, allocating freight by unit count can make the low-cost basics absorb too much cost and the high-value items absorb too little. By-value allocation follows the commercial value of the goods and often lines up well with insurance and some customs-related charges.

By volume or CBM fits ocean LCL shipments and bulky apparel where the carrier's economics are space-driven. Puffer coats, boxed footwear, and structured handbags can consume more container space than their FOB value suggests. If the freight invoice is effectively charging for cube, the landed-cost file should not force the charge through FOB value just because value is available.

By weight is less common for apparel, but it has a place. Dense footwear, leather goods, and heavy outerwear can make weight a better proxy than quantity. For lightweight garments, weight can be noisy because packaging and carton assumptions may matter more than the garment itself.

By quantity works when SKUs are economically similar: one T-shirt program, same fabric family, same carton structure, similar FOB cost, and no meaningful volume differences. Quantity allocation is easy to audit, but it becomes misleading fast when a container mixes high-cost and low-cost styles.

Equal allocation is rare. Use it only for charges that are genuinely line-neutral, such as a fixed documentation fee that policy assigns evenly across invoice lines. It should not be used for freight, duty, or cost components driven by value, space, weight, or origin.

The math should be visible enough for another accountant to audit. If a size row has $1,200 of FOB value inside a $48,000 shipment and the ocean freight pool is $4,200, its freight allocation by value is $1,200 divided by $48,000, multiplied by $4,200, which equals $105. If that row represents 12 units, freight adds $8.75 per unit before duty, MPF, HMF, brokerage, and inland charges.

This is the same allocation discipline behind manufacturer-side landed cost allocation methods, but apparel adds the size-matrix grain. The row being allocated is not the supplier's packed line. It is the expanded style-color-size row that the ERP will actually receive.

Allocate duty from the entry line, not from the shipment average

Duty is not just another shared charge pool in an apparel landed-cost file. Freight can often be allocated across expanded SKU rows by value, volume, weight, or quantity. Duty has already been calculated by classification, origin, preference program, and trade-remedy layer on the customs entry.

The common shortcut is to divide total duty by total entered value and apply the blended rate to every SKU. That is wrong for mixed apparel shipments. A cotton knit top, a synthetic woven dress, a wool coat, and a USMCA-qualifying garment can all sit on the same supplier invoice but carry different duty treatment. A blended rate overcosts some rows, undercosts others, and corrupts per-style margin exactly where apparel wholesalers need precision.

The correct workflow starts with CBP Form 7501. For each entry line, capture the HTSUS classification, entered value, country of origin, duty rate, duty amount, and any Chapter 99 trade-remedy lines. Then map that entry line back to the supplier invoice line or group of lines it covers. That mapping is much easier when the underlying classification fields are already pulled into structured form upstream — fiber composition, knit versus woven construction, shell and lining detail, and country of origin — which is the HTSUS classification extraction step on the apparel commercial invoice itself. Importers who also bring in footwear face a parallel but different field set — outer sole material, upper material, athletic versus non-athletic construction, and FOB tier — covered in the Chapter 64 footwear invoice extraction workflow. Only after that mapping is clear should the duty amount flow down to the expanded style-color-size rows.

Section 301 is the obvious trap. It is origin-specific and line-specific, so Chinese-origin apparel lines can absorb additional duty while similar non-Chinese-origin lines in the same container do not. Temporary Section 122 adds another dated layer: CBP guidance for CSMS 67844987 on temporary Section 122 duties states that heading 9903.03.01 applies a 10 percent additional ad valorem duty to covered imported articles entered on or after February 24, 2026 and through 12:01 a.m. eastern daylight time on July 24, 2026, unless exempt. If that duty is extended, replaced, modified, or allowed to expire, the costing file should change by entry date and HTSUS treatment, not by spreadsheet habit.

Preference programs matter in the other direction. A yarn-forward garment that qualifies under USMCA can have duty reduced to zero, while a visually similar garment from a nonqualifying origin may still carry MFN and trade-remedy duty. A zero-duty row should not absorb duty from nonqualifying rows simply because the shipment total included duty.

The matching keys belong in the workpaper: entry number, entry line number, HTSUS, country of origin, manufacturer style, PO, commercial invoice number, supplier line reference, and entered value. That is why extracting data from customs broker invoices and entry summaries is more than AP automation in this workflow. The broker packet contains the line-level duty structure that protects SKU margin from a blended-rate shortcut.

Reconcile four documents before the receiver posts

The landed-cost workpaper is a reconciliation across documents that were not designed to line up neatly.

The supplier commercial invoice carries the apparel detail: style, color, size matrix or size rows, quantities, FOB unit cost, extended value, currency, origin, invoice number, and PO. It is the source for SKU expansion and product value.

The ocean or air carrier invoice carries freight and shipment charges: base freight, fuel or security surcharges, destination charges, terminal handling, documentation, and sometimes inland delivery. It may reference a bill of lading, container, booking, or airwaybill rather than the supplier invoice number.

The customs broker invoice carries brokerage and passthrough customs fees. It may include MPF, HMF, duty, bond charges, disbursement fees, and entry-related service charges. The broker invoice is also where AP teams often see totals before they have read the entry summary closely.

The CBP Form 7501 entry summary carries the duty logic: entry number, HTSUS lines, country of origin, entered value, duty, MPF, HMF where applicable, and Chapter 99 trade-remedy lines. It is the controlling document for line-level duty allocation.

The matching logic should use more than one key. PO, commercial invoice number, supplier, vessel or flight reference, bill of lading, container number, entry number, and broker reference all help connect the documents. When totals do not tie, the discrepancy usually sits in one of a few places: the broker invoice includes a fee that is not on the entry summary, terminal charges arrive on a separate carrier invoice, MPF is capped at the entry level, or an amended entry changes duty after the first receiver was prepared.

Invoice Data Extraction is useful at this exact upstream point because finance teams can extract supplier invoices, carrier invoices, broker invoices, and CBP entry summaries into one structured working file before the allocation logic runs in Excel, an apparel ERP, or a costing workbook. The product converts uploaded financial documents, including PDFs and images, into structured Excel, CSV, or JSON outputs based on the user's prompt. It should be treated as the extraction and consolidation layer, not as the system that decides accounting policy or posts landed cost into the ERP.

The output of this reconciliation should be a shipment-level workpaper with line-level source references. Every expanded SKU row should trace back to the supplier invoice line that created it, the freight pool that was allocated to it, the broker or entry fee pool it absorbed, and the CBP entry line that determined its duty.

True up late invoices without losing per-SKU cost integrity

The hardest landed-cost problem is not the first estimate. It is the freight invoice or amended broker bill that arrives three weeks after receipt, when half the shipment has already sold.

Start with the incremental cost. If the receiver used an estimated $4,200 ocean freight pool and the actual freight, destination, and terminal total later comes in at $4,680, the true-up pool is $480. Allocate that $480 using the same basis used for the estimate unless the original basis was demonstrably wrong and the company's policy allows correction. If the estimate used FOB value, the true-up should use FOB value. If the estimate used CBM for ocean LCL, the true-up should use CBM.

Then calculate the adjustment at the expanded SKU row. A row representing 2.5 percent of the allocation base absorbs 2.5 percent of the $480 true-up, or $12. If that row has 12 units, the adjustment is $1 per unit. The accounting treatment depends on the ERP and costing policy. Under a standard-cost model, the difference may land in a purchase price variance or landed-cost variance account and be analyzed by style or shipment. Under FIFO or actual-cost layers, the system may adjust remaining inventory layers and push the sold portion through cost of goods sold. The landed-cost workpaper should show the split rather than hiding it in a shipment-level journal entry.

Do not rerun the receiver casually. If inventory has been received, transferred, allocated to wholesale orders, or sold through, reopening the original receipt can create operational fallout that is worse than the cost variance. A controlled true-up file is usually cleaner: same expanded SKU rows, same allocation basis, incremental cost only, and a clear flag for units remaining versus units sold.

The workflow should produce three outputs. The first is an ERP receiving-cost upload or landed-cost import file at style-color-size grain. The second is a per-style margin workpaper for season-buy review, so finance and merchandising can see how freight, duty, and broker costs changed margin by program. The third is a late-invoice reconciliation workpaper that ties estimates to actuals and shows where the adjustment landed in inventory, cost of goods sold, or variance.