Bestsellers in the US, Long-Tail from China: The Hybrid Fulfillment Model for 2026

For years the fulfilment question was framed as a fork in the road: do you ship direct from a China 3PL, or do you hold stock in a US warehouse? Pick one. The trouble is that both answers are wrong on their own, and in 2026, with duty on every parcel and capital expensive, picking one location for your whole catalogue quietly costs you money on either side.

The brands scaling cleanly right now stopped choosing. They split inventory by how fast each SKU sells. This is the hybrid model, why it has become the default for growing D2C and Amazon brands, and exactly how to set it up.

Quick answer: Put your fast-moving bestsellers in a US warehouse for two-to-four-day delivery. Keep slow movers, new-product tests, and non-US demand in China, where storage is cheap and you only ship what sells. Each SKU lives where it is cheapest and fastest, not where a single all-or-nothing decision puts it.

Why "pick one warehouse" is the wrong question

Look at what each all-in choice actually costs you.

Everything in the US. You ship your entire catalogue across the Pacific, pay duty on all of it upfront, and store it on expensive domestic shelves before you know what will sell. Your slow movers and untested SKUs sit there as dead, duty-paid capital. One viral product subsidising fifty that barely move is not a supply chain, it is a cash trap.

Everything in China. Your storage is cheap and your capital stays free, but your bestsellers, the products US customers actually want now, take longer to arrive than a domestic competitor's. On the SKUs where speed converts, you are handing the sale to whoever ships from inside the country.

Splitting by velocity removes both problems at once. We made the full single-location case in our direct-from-China versus US 3PL profit audit; the hybrid model is what you do once you have read it and realised the honest answer is "both, by SKU."

The split logic, in one table

The dividing line is sales velocity and how predictable that velocity is.

Factor	Keep in China	Move to a US warehouse
Sales velocity	Slow / long-tail	Fast / bestseller
Demand predictability	Unproven, seasonal, spiky	Steady, forecastable
Product stage	New launches and tests	Established winners
Target market	Non-US and global buyers	US domestic buyers
Delivery expectation	Standard international	2 to 4 days
Capital exposure	Low, pay to ship what sells	Higher, duty paid upfront

A typical brand ends up with a small number of hero SKUs in the US doing the bulk of revenue at domestic speed, and a long tail of everything else served cost-effectively from Shenzhen.

China as your product lab

The most underrated benefit of the hybrid model is what it does to product testing.

Launching a new SKU into a US warehouse means importing it in bulk, paying duty on it upfront, and renting space for it before a single unit has sold. If it flops, you own a pile of duty-paid stock you cannot move. That risk makes brands timid about testing, which is the opposite of what a growing catalogue needs.

Fulfil the launch direct from China instead and the risk inverts. No bulk import, no upfront duty, no committed shelf space. You pay to ship each unit only when it actually sells. Let the SKU prove itself in market, and once demand is steady and forecastable, graduate it into the US warehouse as a new bestseller. China becomes a low-risk lab where you can run far more product bets than a US-only operation could afford, then promote only the winners.

Feeding the US leg without the storage bill

Once a SKU earns its place in the US, you still control how you stock it. You do not have to ship a year of inventory at once.

The smart pattern is to import proven bestsellers from China in right-sized, frequent batches. For US-bound bulk, ocean freight on a delivered, duty-paid basis runs roughly $110 to $180 per cubic metre for LCL as a 2026 guide, with air freight around $4.50 to $8.50 per kilogram when you need speed for a launch or an urgent restock. Treat sea as the default for replenishing volume and air as the exception. Because we consolidate your suppliers in Shenzhen first, several factories' goods ship together on one entry, so duty is paid once and your per-unit freight drops. The same consolidation feeds drip-fed FBA replenishment if Amazon is your US channel, keeping you under Amazon's storage-fee thresholds.

What this looks like running day to day

• China holds your long tail, your launches, your seasonal lines, and your stock for the UK, EU, Australia and the rest of the world, all shipped DDP so buyers see no surprise duty.
• The US holds a tight set of proven bestsellers for two-to-four-day domestic delivery and easy returns.
• Shenzhen is the hub that receives from all your suppliers, runs quality control, and feeds both legs: direct-to-consumer for the long tail, bulk freight to top up the US winners.
• You shift SKUs between legs as the data moves. Winners graduate to the US; faders drop back to China-only. The split is never static.

The bottom line

The 2026 winners are not the brands that found the one perfect warehouse. There isn't one. They are the brands that put each SKU where it is cheapest and fastest to serve, used China as a cheap, capital-light lab and global shipper, and reserved expensive US shelf space for the proven few that earn it. We run the China side of that model, and the freight that feeds the US side, from one Shenzhen operation. Send us your top SKUs and we'll model the split with you.