In the news it is a trend; on the broker's desk it is one more document stack per origin. What each new origin replicates — classification, origin files, local export compliance, exam-risk history — why only substantial transformation changes origin and relabeling never does, Taiwan's 15% non-stacking terms in the comparison math, and duty-paid US inventory as the buffer against policy swings.
The December question — can we get into Amazon next week? — was answered in October. The three-stage Q4 jam and its propagation effect, the outside warehouse as FBA overflow buffer that preserves the option to ship, the backward-planned stocking timeline with buffer stacked at the latest segments, and the contingency list with pre-agreed plan-B triggers.
Sellers using outsourced warehouses do not need to buy a WMS — they need the window the warehouse's WMS opens to them. The thirty-second WMS-vs-ERP split, the three things a 3PL owes you (real-time inventory, timestamped records, API integration), the questions to ask about APIs without reading code, and the data-capability checklist that ends with the quiet test: can I export my history if I leave?
Same container, same dock, very different fates: one batch goes to a rack and waits, the other sorts straight to an outbound truck. What cross-docking actually is, the time-and-fee ledger against store-and-ship, the four cargo profiles that fit and three that do not, and the one-question decision line: was the destination fixed the day the goods sailed?
Checklists get memorized; blank fields cause the holds. The baseline document set, the additions by cargo profile, and the four fields with the highest hit rates on the Taiwan-US lane — shipping marks, origin statements, the HS six digits, currency and trade terms — plus the day-by-day cost of one wrong field computed.
Two factories, same box, an 8 CBM difference — stuffing is where ocean freight saves money without spending any. Nominal versus real loadable volume for the three workhorse types, the pallet-versus-floor-loading trade, why dense cargo weighs out before it cubes out, and the US highway weight limits that bind before the ship does.
The quiet way to lose money is saving on ocean freight and paying it back with two thousand miles of trucking. What each of the five US gateways covers, their transit ranges and congestion temperaments from Taiwan, and the decision logic that starts from where the buyer sits — not from "everything goes through LA."
Ocean transit is a range by nature — quoting a fixed date is the exporter's classic mistake. Typical port-to-port ranges from Taiwan to the five US gateways, what direct versus transshipped routings change, the customs and inland days door-to-door adds, and why peak-season lead times quote off the top of the range.
A warehouse holding food is itself an FDA-regulated food facility — the "we only store it" exemption does not exist. The four gates (registration, continuous temperature records, pest control, lot traceability), matching four temperature tiers to product labels instead of assuming colder is better, and the 10-point checklist for a site visit.
Returns never appear as one line item — they hide across freight, labor, storage and write-offs, quietly rivaling the category margin. Why return rates are category physics, the four payments in every return, the three levers that lower them (size charts, local refurbishment, policy design), and why free returns is a price input rather than a slogan.
