Roughly $30 billion of reciprocal imports in each direction sit on newly published U.S.-China tariff relief lists — 77 categories of Chinese goods on the U.S. side and 1,619 categories of American products on China's — but none of the lower rates are in effect yet. Both governments still must complete domestic implementation procedures, and no comprehensive effective date has been announced.

That gap between announcement and application defines the operational challenge for importers and exporters. Purchase orders, production schedules and ocean transit times keep moving while trade policy catches up. A shipment can leave one Chinese port under one tariff assumption and clear U.S. customs under another.

China has said more than 90% of the products covered by the arrangement could shed the additional tariffs imposed during the trade conflict and return to standard most-favored-nation rates. But the deal is far from normalization. Semiconductors, batteries and electric vehicles remain outside the agreement, and the overall tariff structure stays considerably more restrictive than before the conflict escalated.

From country-level to SKU-level decisions

For years, companies treated China tariffs as a country-level sourcing problem: stay in China or move to Vietnam, Thailand, India or Mexico. The new lists show why that framework no longer works. Two products from the same supplier, moving through the same Chinese port to the same distribution center, can now face materially different landed-cost structures based purely on classification.

The operative questions sit at SKU and component level. Which product qualifies for relief? Which component stays dutiable? What country of origin applies after assembly? Will the shipment enter before or after the new rate takes effect? Those are not customs questions alone — they cut across sourcing, transportation, inventory planning and network design.

The list composition reinforces the granularity. The U.S. list covers familiar containerized consumer goods: toys, blankets, tableware, artificial flowers, child safety seats, holiday decorations. China's list spans agricultural products, meat, seafood, dairy, timber, personal-care products, medical equipment and coal. On a trade-policy spreadsheet these look unrelated. In logistics terms they map into very different networks — transpacific container routes for consumer goods; elevators, rail, barges, cold storage and refrigerated export facilities for agriculture; rail and bulk-terminal capacity for coal; premium ocean and airfreight for medical equipment.

A tariff change therefore does more than alter the duty at the border. It can change which supplier wins the order, which mode carries the freight, how much inventory a company holds, and which ports or inland corridors absorb volume. Trade policy is increasingly functioning as a freight-demand signal: a decision in Washington or Beijing becomes container bookings from Shenzhen, intermodal volume in Chicago, grain movements on the Mississippi River or reefer capacity at a Gulf Coast port.

The data behind the diversification

Container data shows how far sourcing diversification has already progressed. U.S. containerized imports reached approximately 2.6 million TEUs in August, among the highest monthly totals on record. China remained the largest origin at roughly 884,000 TEUs, but its share of total imports continued to edge lower while volumes from Vietnam, Thailand and Indonesia increased. September volumes stayed elevated as retailers positioned holiday merchandise.

Much of the immediate freight response has already happened. Purchase orders were placed months ago, containers are on the water and inventory is flowing through distribution centers. Tariff relief announced now cannot unwind those decisions.

The more consequential effects may appear in 2027 sourcing negotiations, when procurement organizations must decide whether production moved out of China should stay put, whether Chinese suppliers should regain share in specific categories, and whether dual-sourcing arrangements should be adjusted rather than abandoned. In many cases the answer will differ by product, not by country.

Consider a retailer that shifted a household product from China to Vietnam because elevated tariffs made the Chinese supplier uneconomic. If that differential narrows substantially, the original supplier may become attractive again on manufacturing scale, established tooling, supplier ecosystems, production efficiency and port connectivity. The point is not that the retailer should move the business back — diversification may have delivered resilience worth keeping — but that the underlying assumptions deserve a fresh test.

Agriculture feels it through the network

China's relief list includes wheat, corn, sorghum, meat, seafood, dairy, soybean oil and soybean meal. Commercial soybeans themselves remain outside the arrangement despite their historic weight in U.S.-China agricultural trade. Additional Chinese demand for U.S. corn, wheat or sorghum would move quickly into grain elevators, barge movements, rail capacity, export terminals and bulk shipping. More meat and dairy exports would require cold-storage infrastructure and reefer containers. Coal purchases would generate rail movements and bulk-vessel demand.

For shippers and forwarders, the practical response is to recalculate, not to declare China back or China-plus-one dead. Product classifications and landed costs need revisiting. Open purchase orders need mapping against tariff implementation dates. Chinese suppliers need benchmarking against alternatives using current freight, lead-time, inventory and risk assumptions. Companies should also separate sourcing shifts that genuinely built resilience from those that were pure tariff arbitrage — the former may hold strategic value even after the tariff advantage disappears; the latter deserve another look.

Network design itself is changing character. Supply chains were once modeled on assumptions meant to hold for years. Today a tariff revision, sanctions change or export control can shift the economics of a sourcing lane within months, turning network design from a periodic strategic exercise into a continuous operating discipline.

The result is likely to be a supply chain still deeply connected to China but structurally less dependent on any single manufacturing geography — with selective sourcing decisions made product by product, supplier by supplier and lane by lane.