A SKU placement matrix gives each sellable SKU one auditable recommendation: keep it in China, pre-stock it in the US, use a capped split, or test it from China first. Score the operating evidence behind that choice, then apply hard stops before moving inventory. Bondjet helps make the China-side evidence usable through receiving, inspection, SKU management, protective packaging, inventory handling, and international dispatch.
A warehouse decision is expensive to reverse once stock has been received, counted, stored, and promoted in a new market. The useful question is not, “Which warehouse is best?” It is, “Which location can serve this SKU's current demand with an acceptable margin and a controlled fallback?” This guide provides a downloadable-ready SKU placement table, a scoring method, and the approval rules that keep a score from becoming an automatic transfer order.
Key Takeaways
- Use one row per sellable SKU or variant, with source metrics, assumptions, recommended node, confidence, owner, and review date on the same record.
- Score velocity, contribution margin, weight or cube, seasonality, return risk, stockout cost, and replenishment lead time using weights set by your business.
- Apply hard stops for prohibited goods, unstable packaging, unresolved returns, missing demand evidence, or an unconfirmed local operating path before any score is used.
- Map the result to four actions: China only, US only, capped split, or a controlled China-direct test.
- Bondjet can support the China-side controls behind the matrix, while a US warehouse, local fulfillment, and returns program must be confirmed separately for the proposed arrangement.
What Is a SKU Placement Matrix?
A SKU placement matrix is a decision worksheet that compares each product's commercial and operational evidence against the requirements of each inventory node. It does not promise a universal threshold or replace judgment. It makes the assumptions visible before inventory is scattered across warehouses.
The matrix works best at the SKU-variant level. One color, size, bundle, or packaging revision can have different demand, dimensions, margin, and return behavior from the product family around it. Grouping them too early can make a fast-moving variant appear to justify stock for a slow-moving one.
For China-versus-US placement, the matrix should record the service promise as well as cost. A US node may make sense for a stable hero SKU with a defined local fulfillment path. A long-tail or newly launched SKU may be safer in China until its demand, packaging, and service results are repeatable. Bondjet 邦达捷 can provide the origin-side controls needed to make that assessment more reliable, including receiving records, SKU checks, photos where requested, protective packing, and international dispatch.
[Image suggestion: a four-quadrant placement matrix with China only, US only, capped split, and China-direct test as the four outputs.]
Start with evidence, not a catalog-wide rule
“Put bestsellers in the US” is a useful hypothesis, not a rule. A bestseller with thin margin, high cube, unpredictable returns, or a six-week replenishment cycle can still create costly local inventory exposure. Likewise, a slower item may deserve a small local allocation when a committed sales channel needs a specific delivery promise and the commercial case is documented.
Use common definitions before comparing teams' numbers. The ASCM APICS Dictionary is a helpful reference for terms such as lead time, available inventory, and safety stock. Your matrix should then state the business-specific definition used for each field.
Build Your SKU Warehouse Placement Template
The table below is designed to be copied into a spreadsheet or CSV. Keep raw data in its source systems, but freeze the values used for each decision. A score is only useful when another owner can see the date, source, and assumptions behind it.
| Column | What to enter | Why it belongs in the decision |
|---|---|---|
| SKU / variant | Unique sellable item and packaging version | Prevents one variant from masking another's behavior |
| Channel and market | Store, marketplace, wholesale account, and destination market | Shows where demand and service obligations actually originate |
| Trailing units | Weekly or monthly units for a stated lookback period | Supplies the velocity input without hiding the data cut |
| Contribution margin | Revenue less product, handling, packaging, shipping, support, and expected returns | Tests whether the SKU can absorb a second inventory model |
| Weight and cube | Packed weight and dimensions, plus the chargeable method used | Captures transfer, storage, and carrier exposure |
| Seasonality | Peak period, event dependency, and expected demand volatility | Stops a short spike from becoming a broad stock commitment |
| Return and damage rate | Reason-coded returns, damage, replacements, and restockability | Shows risk that a sales-only score misses |
| Replenishment lead time | Supplier preparation through destination availability | Shows whether local stock can be replenished before it runs out |
| Target service | Customer promise, channel rule, or cutoff requirement | Makes the placement answer a service question as well as a cost question |
| Proposed node | China only, US only, capped split, or China-direct test | Turns the evaluation into an actionable recommendation |
| Confidence / owner / review | High, medium, or low confidence; named approver; next review date | Keeps exceptions and uncertainty visible |
Downloadable SKU placement table
Copy this header row into a CSV or XLSX worksheet. Add a source-data tab rather than overwriting values after a decision is made.
sku,variant,packaging_version,channel,market,trailing_units,lookback_days,contribution_margin_pct,packed_weight_kg,packed_cube_m3,seasonality_note,return_damage_rate_pct,replenishment_lead_days,target_service,velocity_score,margin_score,cube_score,seasonality_score,returns_score,lead_time_score,hard_stop,recommended_node,allocation_cap,confidence,exception_owner,approval_owner,next_review_date,decision_note
The result is downloadable-ready because each cell can be reviewed, filtered, and exported. It is not a generic calculator. Add columns when a product needs them, such as hazardous-goods eligibility, minimum order quantity, inspection instruction, bundle-component availability, or a named return disposition path.
Score the SKU Placement Matrix Without Pretending It Is Universal
Assign a 1-to-5 score to each factor, then multiply it by a weight selected by the business. A 5 should mean “strongly supports the proposed node,” while a 1 means “creates material exposure.” Set the definitions before seeing the answer.
| Factor | Example interpretation for a higher score | Why the weight may change |
|---|---|---|
| Velocity | Stable demand over the selected period, not one campaign week | A brand with high inbound cost may weight predictability more heavily |
| Contribution margin | Margin remains acceptable after node-specific handling, storage, delivery, and return reserve | A low-margin catalog may need a stricter gate |
| Weight / cube | Compact, economical to transfer and store for the proposed service level | Bulky items can make local stock expensive even when demand is strong |
| Seasonality | Demand is predictable and stock can exit before the season ends | Event-led products need a shorter review cycle |
| Returns / damage | Return reasons are manageable and packaging controls are proven | Condition-sensitive products may weight this above parcel cost |
| Stockout cost | A missed sale or channel penalty is material and documented | A promised delivery window can raise the value of local availability |
| Lead time | Replenishment can arrive before projected cover is exhausted | Slow or variable supply favors smaller allocations and buffers |
One simple calculation is:
Placement score = sum of (factor score x factor weight)
Use the result as a discussion input, not a transfer trigger. A high score can support a US allocation only after the physical route, receiving process, inventory ownership, fee model, cutoff, system sync, exception contact, and returns path have been confirmed. A lower score may support China-direct fulfillment if the customer promise, route, and customs arrangement fit the actual order.
For imports into the United States, warehouse planning does not remove importer responsibilities. Review the shipment-specific arrangement with appropriate specialists; U.S. Customs and Border Protection's importing guidance is a practical starting point for understanding the import process.
Use a hypothetical sample row
The example below illustrates the logic only. It is not a recommended threshold, price, or service promise for every seller.
| SKU | Trailing units | Margin | Weight / cube | Seasonality | Returns | Lead time | Hard stop | Recommendation |
|---|---|---|---|---|---|---|---|---|
| FIG-ALPHA-01 | 240 units / 90 days | 42% contribution | 1.1 kg / 0.012 m3 | Stable outside a short holiday lift | 2.5%, with documented box-condition checks | 28 days to US availability | None; local receiving and returns process confirmed | Capped split: move four weeks of approved cover, then review |
The row does not say that every collectible should be pre-stocked. It says this hypothetical SKU has enough documented evidence for a limited test. For a high-value or fragile item, Bondjet's collectible fulfillment workflow can help evaluate China-side inspection, SKU verification, protective packaging, and dispatch records before a placement decision is scaled.
Add Hard Stops Before a China or US Warehouse Decision
Hard stops are non-negotiable conditions. They sit above the weighted score because a mathematically attractive result cannot repair an unavailable service or an unresolved product risk.
| Hard stop | Why the score cannot override it | Required next action |
|---|---|---|
| Product or route is ineligible | Goods may not be accepted on the intended service | Verify eligibility and select a lawful, confirmed route |
| Packaging is unstable | Damage and customer disputes can increase with added handoffs | Revise pack-out and prove it with controlled shipments |
| Returns are unresolved | A local allocation can create stranded or unaccounted inventory | Name the return address, inspection standard, owner, and disposition rules |
| Demand evidence is distorted | A launch promotion, stockout, or wholesale order can misstate ordinary demand | Tag the event and extend the evidence window |
| Local operation is unconfirmed | A claimed node may lack receiving, systems, cutoff, or service coverage | Obtain written scope, fees, process, and exception contacts |
| SKU record is incomplete | Wrong variants, accessories, or dimensions corrupt the score | Reconcile the SKU master and physical inventory first |
Bondjet's documented strength is China-side fulfillment: receiving, inspection, SKU management, warehousing, packaging, and international dispatch. Do not represent a US warehouse, domestic pick-and-pack SLA, or returns service as included unless the specific location and terms have been confirmed. That boundary is what makes the matrix credible.
Map the Result to Four Inventory Placement Actions
After the score and hard-stop review, choose one action. Record the allocation cap and the next decision date in the same row.
China only
Use China only when the SKU has volatile or low demand, many variants, uncertain returns, high risk of aging, or no confirmed local operation. This preserves flexibility while the team gathers more evidence. Bondjet can keep the origin-side workflow controlled through SKU records, inspection options, packaging instructions, and dispatch coordination.
US only
Use US only when demand is stable, the margin supports local handling and storage, service requirements are clear, replenishment is workable, and the destination operator is fully confirmed. Even then, review the decision at a fixed cadence because demand and storage exposure change.
Capped split
Use a capped split when a hero SKU benefits from local availability but the business wants to limit aged stock. Set the cap in weeks of cover or units, not a vague “small amount.” Keep a China-direct fallback only where the customer promise allows it.
Test from China
Use a controlled China-direct test for new SKUs, revised products, uncertain packaging, or demand with a short history. Define the test quantity, customer promise, evidence window, and promotion gate first. The new SKU fulfillment test guide explains how to review demand, margin, packaging, service, and replenishment evidence before moving inventory closer to customers.
[Image suggestion: a decision tree showing hard-stop check, weighted score, then China only, US only, capped split, or China-direct test.]
Review Confidence and Exceptions Before Every Move
The matrix is a living operating record, not a one-time spreadsheet. A high-confidence recommendation has current data, stable definitions, confirmed capabilities, and no material exceptions. A low-confidence recommendation should not quietly become a transfer because someone needs an answer quickly.
Use an exception owner for every uncertainty: late supplier output, a packaging change, a return spike, a channel promotion, a product revision, an inventory discrepancy, or an unconfirmed destination process. The owner should state the action, deadline, and condition that allows the score to be reconsidered.
Run a full SKU review monthly and use an exception queue between reviews. The monthly inventory rebalancing rules provide a useful cadence for reconciling sellable inventory, checking cover against full lead time, approving transfers, and documenting the next review. Before moving a larger allocation, replay historical demand under China-only, US-stock, and hybrid assumptions with fulfillment scenario backtesting. Compare stockouts, delivery promise, transfer timing, storage, returns, and contribution margin together.
FAQ: SKU Placement Matrix
How do I decide whether a SKU belongs in a China or US warehouse?
Start with one SKU-level row that records demand, contribution margin, packed weight or cube, seasonality, returns, lead time, and required service. Apply hard stops first, then choose China only, US only, a capped split, or a China-direct test. Confirm the actual destination operation before transferring stock.
What is the most important field in a SKU warehouse placement template?
There is no single most important field because a warehouse decision has commercial and operational consequences. The most useful field is often the decision note, because it connects the recommendation to the data cut, assumptions, confidence, owner, and next review date. Without that record, a score cannot be audited.
Can a high score automatically approve US inventory?
No. A high score only indicates that the documented evidence supports further review. It cannot override prohibited-goods rules, packaging problems, unresolved returns, incomplete SKU data, or an unavailable local service. A named owner must confirm the transfer quantity, operating path, and fallback before inventory moves.
Conclusion: Make Every SKU Placement Decision Explainable
A useful SKU placement matrix does not tell every seller to stock more products in the US. It makes the trade-offs visible: demand quality, contribution margin, cube, seasonality, returns, stockout cost, lead time, and service requirements. Add hard stops, name an exception owner, cap the first move, and set a review date so the decision stays reversible where possible.
For high-value, fragile, or complex-SKU products, start with the origin-side details that often distort a warehouse decision: accurate SKU records, condition checks, packaging instructions, inventory visibility, and dispatch evidence. Contact Bondjet with your SKU list, target markets, current demand data, packaging requirements, and proposed service promise. Bondjet can help assess the China-side fulfillment controls behind the matrix, then support a more disciplined China or US warehouse decision without overstating unconfirmed destination services.