China 3PL vs US 3PL: The Real Cost and Speed Math

The decision between a China 3PL and a US 3PL is usually argued on delivery speed. That is the wrong axis. The number that actually moves is where your parcel enters the USPS network, because USPS prices and routes by distance zone from the point of induction, not from where the goods were made. The same 1 lb parcel to Manhattan costs $7.61 inducted in New York and $10.67 inducted in Los Angeles, a 40% difference on identical goods (USPS Notice 123, commercial prices, effective 12 July 2026). This post works the whole comparison with published rate cards, USPS's own rulebook, and the duty stack as it stands in August 2026.

A warning about most content on this topic: much of it is now wrong. De minimis ended in June 2026, so any article claiming China-direct avoids duty is out of date. The IEEPA tariffs were struck down by the Supreme Court in February 2026, so any article quoting them is out of date too. Everything below is dated and sourced.

The zone rule almost nobody applies correctly

USPS prices zoned mail on distance between two points. The question that decides your shipping bill is which two.

USPS answers it three times, in three different parts of its own rulebook, and the answer is always the same: the origin is the US post office where the parcel is deposited, not any foreign origin.

• The geometry, from the Domestic Mail Manual 608.9.1: distance is measured "from the centroid of the 3-digit ZIP Code area serving the origin Post Office to the centroid of the 3-digit ZIP Code area serving the destination Post Office."
• The operational rule, from DMM 604.4.6.3: postage is calculated "taking into account the weight and the zone from the ZIP Code of the office where the mailpiece is deposited to the ZIP Code of the destination address."
• The enforcement, from DMM 705.20.5.5: USPS audits zone claims against the physical entry location and assesses the difference when a piece claims a lower entry zone than the one calculated from where the container was actually entered. The error threshold is 0.01%, which is about as close to zero tolerance as USPS gets.

That last one matters most, because it is USPS putting money behind the definition. They verify against the physical entry point, so the physical entry point is the fact that counts.

Why an imported parcel can be treated this way at all is the part that gets missed. Under International Mail Manual 621, Global Direct Entry wholesalers "tender international inbound shipments and parcels to the Postal Service after the U.S. Customs and Border Protection has cleared the items." IMM 622.1 then requires those items to be prepared to domestic standards and paid for as domestic products. By the time USPS touches them they are domestic mailpieces carrying domestic postage, and ordinary zone rules apply from wherever they were handed over.

Here is the contrast worth holding on to. IMM 790 restricts inbound international mail: it "must be entered through one of the five international service centers located in the 48 contiguous states." Five fixed gateways. That constraint binds the postal channel. It does not bind customs-cleared parcels tendered as domestic mail, which is precisely why the commercial channel can induct near the customer and the postal channel cannot.

What induction point is worth, in dollars and days

Same parcel. Same destination. Two induction points.

Seven zones separate those two rows. The parcel, the weight and the customer are identical.

Now the full ladder, because the penalty is not linear. These are USPS published commercial prices, which are a ceiling rather than a floor: high-volume shippers negotiate below them, and USPS has confirmed that negotiated commercial customers are unaffected by its published-rate changes.

Two things to notice. The 1 lb spread from Zone 2 to Zone 8 is 38.9%. And the sharpest single step is Zone 4 to Zone 5, where 2 lb jumps 16.9% for crossing one line on a map.

Three honesty notes on that table, because they are the kind of thing a sharp prospect will test you on:

1. The day bands are derived, not published. USPS does not publish a zone-to-days table. These come from joining USPS's own 3D Base Service Standard Directory to zone lookups across 190 origin-destination pairs. The values are USPS's, the arithmetic is ours.
2. Service standards are goals, not guarantees. USPS defines a service standard as "a stated goal for service achievement." Ground Advantage carries no money-back guarantee.
3. Zone and days are computed separately. USPS derives service standards from facility drive-time bands, not from zones. They correlate tightly because both track distance, but "Zone 2 means 2 days" is a correlation dressed up as a rule. The honest statement is that near-destination induction lands in the 2 to 3 day band and single-West-Coast induction lands in the 4 to 5 day band.

Also note every USPS price quoted here includes a temporary 8% transportation surcharge in force from 26 April 2026 until 17 January 2027, which USPS describes as "a necessary bridge to a permanent mechanism."

The conclusion is narrower than most vendors will tell you, and it cuts both ways. Zone 8 is open-ended, so a parcel inducted at a West Coast gateway pays the same Zone 8 to Philadelphia that a Los Angeles warehouse pays. Crossing an ocean earns you nothing by itself. This is not an argument that China beats America. It is an argument that a single-node fulfillment network is expensive wherever that node sits.

Which is why the question to ask any cross-border provider is not "how fast do you ship" but "where do my parcels enter the domestic network." If the answer is one coast, you have bought a Los Angeles warehouse with a longer first leg.

Our own US-bound parcels clear customs and enter the domestic network on both coasts, so an order going to Philadelphia is inducted in the New York metro rather than trucked across the country from California. On the published commercial prices above, that is what it does to the same 1 lb parcel:

(Zones derived from the DMM 608.9.4.2 distance radii between induction point and destination. Prices are USPS published commercial Ground Advantage, Notice 123, effective 12 July 2026.)

The whole Northeast corridor, which is the densest consumer market in the United States, lands in the near zones and the two-day band. From a single Californian node it is Zone 8 and the four-to-five-day band, every parcel, every day.

What a US 3PL actually costs, itemised

Published rate cards are rare in this industry. The most useful public dataset is The Fulfillment Advisor's annual survey, which covers "over 600 warehouses each year, covering 62 questions." Its 2025 averages:

A worked month at 1,000 orders, using those averages, one pallet of cover, before postage:

Then postage on top. Using the published commercial 1 lb rates above and an illustrative national split of 30% of orders in near zones and 70% in far zones from a single Californian node:

(Order mix is illustrative. Zone prices and the survey averages are published figures.)

That annual difference is not a discount anyone negotiated. It is geography, priced by USPS, on the same parcels to the same customers.

The squeeze that unit rates hide

This is the finding that surprised me most, and it is the one no competitor article reports.

Between the 2024 and 2025 surveys, headline 3PL rates went flat or down. Pick and pack moved from $3.18 to $3.20. Receiving fell from $12.91 to $10.52. Read only those and you would conclude US fulfillment got cheaper.

Now read the terms:

The unit price stayed still while the commitment, the lock-in and the loss rate all moved against the customer. A brand comparing quotes on pick-and-pack alone is reading the one column that did not change.

Note the surcharge line honestly: long-term storage penalties became roughly twice as common while becoming smaller individually. Both halves are true and quoting only one of them tells a false story.

The peak season trap

Storage is quoted annually and charged seasonally. Amazon publishes the clearest example, in its own Multichannel Fulfillment rate card dated 1 June 2026:

That is 3.08 times the base rate, arriving in exactly the quarter when you are holding the most inventory. Amazon also added a 3.5% fuel and logistics surcharge to US fulfillment fees in 2026, described in its own words as being applied "consistent with other major carriers."

Any annual storage model built on the January rate understates Q4 by a factor of three. If you are planning peak, our Q4 cutoff calendar works the dates backwards from Black Friday.

One comparison to avoid, because it is the most common error in this space: Amazon's MCF fulfillment fee of $7.34 for a single-unit small standard order is all-in, including postage. A 3PL's $3.20 pick and pack is not. Setting those two side by side manufactures a gap that does not exist.

Duty is paid either way. The timing is what differs

This is where most China fulfillment content is now simply out of date, and where an honest answer is more persuasive than a clever one.

De minimis is gone. CBP suspended the exemption indefinitely for non-postal shipments effective 24 June 2026 and for mail effective 24 July 2026 (91 FR 37789). The context is scale: CBP recorded over 1.36 billion de minimis shipments in FY2024, against 139 million in 2015. Entry Type 86 is withdrawn. A parcel arriving from China today pays duty. Anyone telling you otherwise is quoting a dead rule.

The stack on a Chinese-origin cotton T-shirt, as of August 2026:

Two corrections to what you may have read elsewhere. The IEEPA tariffs are no longer collected: the Supreme Court held in February 2026 that IEEPA does not authorise tariffs, Executive Order 14389 ended them, and CBP stopped collection at 12:00 a.m. ET on 24 February 2026. The Section 122 surcharge that replaced them ran 150 days from 24 February and has now lapsed. Smartphones and laptops are also exempt from the forced-labor duty under HTSUS note 52(b).

Now the part that actually matters for the comparison. Both routes pay that 36.5%. What changes is when, and on how much.

Import in bulk and you pay duty on your forecast. Ship per order and you pay duty on your sales. On a 70% sell-through, that is $1,314 of duty on this example that the per-order route never spends, and the container route never gets back. Same tariff schedule, different exposure to being wrong about demand. The capital side of this is worked in full in our cash-flow math post.

The per-entry cliff

One more number decides whether per-parcel fulfillment is viable at all, and almost nobody discusses it.

The Merchandise Processing Fee is charged per entry, not per dollar:

(FY2026 figures, in force to 30 September 2026. From 1 October 2026 the formal minimum rises to $34.58 and the maximum to $670.86.)

Read the first row again. If every parcel required its own formal entry, 1,000 parcels a month would carry $33,580 in MPF alone. That is the cliff. It is why the entry mechanics behind a China-direct operation matter more than the headline shipping rate, and why "we ship direct from China" is an incomplete answer without knowing how the goods clear.

The informal entry threshold is $2,500 (19 CFR 143.21), a bond is now required for postal informal entries under the new 19 CFR 145.15, and CBP's Entry Type 13 electronic mail process begins testing on 22 September 2026. This area is moving quarterly. Ask any fulfillment partner, us included, to explain their clearance path in specifics before you commit volume.

When a US 3PL is the right answer

An honest comparison names the cases it loses, and there are real ones.

Choose a US 3PL when:

• You need one or two day delivery as a product promise. Nothing crossing an ocean competes with a domestic node for that, and Shopify's own Shop Promise badge sets its bar at "five or fewer calendar days" domestically.
• Your demand is stable and predictable. The whole cost of inventory is the risk of being wrong. If you are not wrong, that cost is small and bulk freight economics win.
• Your product is heavy, bulky or low value. Per-parcel international line-haul punishes weight far harder than sea freight does.
• You are already holding US stock profitably. Do not fix what is working.

Choose China-direct when:

• You are testing products and cannot afford to be wrong about quantities.
• Your catalogue is wide and your per-SKU velocity is low, where storage and minimums bite hardest.
• Your cash is better used on inventory turns or ad spend than sitting in a warehouse in the Inland Empire.
• Your customers accept a 3 to 10 day delivery window, which most do when shipping is free and the date is accurate.

On that last point, the evidence is less flattering to speed than the industry likes to admit. Baymard Institute's meta-analysis of 50 studies puts documented cart abandonment at 70.22%, with extra costs including shipping cited by 40% of abandoners and delivery being too slow cited by 20%. Delivery cost drives abandonment about twice as hard as delivery speed does. The defensible reading is that cost drives abandonment, unreliability drives churn, and raw speed is a weaker lever than most vendors selling speed will tell you.

How to run this math on your own numbers

Averages are for articles. Your SKU is what matters.

1. Weigh your product, properly, packed. Weight decides your shipping line and it is the variable with the most power over your margin.
2. Find your true zone mix. Pull last quarter's orders, map destination ZIPs, and see what share sits in Zones 5 to 8 from your current node. That single number tells you whether induction point is worth anything to you.
3. Get the whole 3PL quote, not the pick fee. Minimum monthly spend, contract length, long-term storage triggers, receiving, and the shrinkage allowance. The unit rate is the column that did not move.
4. Model duty on forecast versus duty on sales, at your realistic sell-through. If you sell through 100% of everything you buy, the container wins. If you do not, the gap is the duty and freight on the units that never sold.
5. Price a real product both ways. Our Dropshipping Calculator returns landed cost and margin per unit against your destination market, free on the $0 plan.

We ship China-direct from our own Shenzhen warehouse, dispatch in under 24 hours, induct on both US coasts so your east-coast customers are not served from California, and deliver in 3 to 10 days on the local carrier your customer already trusts, with 99.8% delivery accuracy. What we do not do is hold your stock in America, so if this article has convinced you that a US node is right for your product, believe the math rather than the vendor. If it has convinced you the opposite, connect your store and price a live SKU, or read the ShipBob comparison for the same argument aimed at one competitor.