You paid your factory in January. The goods crossed the ocean in February, cleared customs in March, and sat in a warehouse through April. The customer paid you in May.
For four months, your cash sat in boxes you hadn't sold yet, money you could've been using to grow.
That's a cash-flow problem, and brands try all kinds of solutions to close it, from better factory terms to bank financing. But lately one approach keeps coming up: direct fulfillment, straight from the factory to your customer's door.
Here's how it works.
You keep your inventory at or near the factory and fly each order out as it sells. You pay for the goods as they go, not months ahead.
On paper, it's a no-brainer. You free up a pile of cash, hold less inventory, and grow faster.
So why isn't everyone doing it? Is it only for certain products? And how do you know if it's right for you?
To find out, I talked to two people who see it from opposite sides. Izzy Rosenzweig founded Portless, a 3PL built entirely on this model and used by hundreds of brands. Devin Knight is VP of Transport and Planning at ShipMonk, a major 3PL that deliberately doesn't do it.
I also dug into how Quince, which built its whole business on this model, runs its supply chain.
Here's what I found.
Table of contents
First, what does "direct from China" actually mean?
If you've only heard the term in passing, here's how the model works.
In the traditional setup, you manufacture a big batch overseas, ship it across the ocean in a container, pay duty on the whole shipment when it lands, and store it in a US warehouse until customers order. You've paid for everything months before you sell a thing.
Direct fulfillment changes one part of that: where your inventory sits. Instead of crossing the ocean in bulk, your stock stays at a facility near the factory. Nothing moves until a customer orders.
Then, the order is:
Picked and packed near the factory.
Flown to the US that day or the next.
Dropped at a regional "injection point," an airport where it enters the domestic mail stream.
Delivered the last mile by a carrier, like USPS.
Your customer can't tell the difference.
They get a tracking number, a delivery from a familiar carrier like USPS, and their order in five to eight days, the same as anything else they buy online. Izzy Rosenzweig at Portless calls it a "100% local experience."
The changes are all on your side. Here are the two big ones:
Your inventory sits at the factory, not in a warehouse near your customers. You stop pre-positioning months of stock in your destination markets.
You pay per order, as it sells, instead of all at once before anything sells.
The second one is the one to watch. It's where the whole advantage comes from.
What you're really deciding: cash vs. control
On the surface, direct fulfillment looks like a logistics question: where to put a warehouse, how fast the box arrives, who runs the last mile.
But the more useful frame is financial.
What you're really choosing is how to fund your inventory, and where the goods sit follows from that.
Framed that way, the decision comes down to understanding three things:
What you gain: cash
Almost nothing leaves your account until the customer has paid you.
What you give up: control
You take on harder returns, slower delivery, and the question of whether your brand can wear "ships from China."
What qualifies: light, well-margined products
The model only works if your product is light (under ~3.5 lb), carries enough margin to absorb the costs (above ~40%), and isn't dirt cheap (AOV over ~$25).
More on each below.
What you gain: cash
Improving cash flow is the whole reason the model exists, so let's break it down.
In the traditional model, your money goes out long before it comes back:
You pay the factory for a full production run.
You pay to ship it across the ocean, 45 to 60 days in transit.
You pay duty on the entire shipment the day it lands.
Then it sits in a warehouse, waiting to sell.
By the time it all plays out, the median public DTC brand is carrying about 133 days of inventory (Eightx's read of SEC filings), roughly four months of goods on the shelf before a customer ever pays.
On $900K of annual goods, that works out to about $330K of your cash tied up in stock you haven't sold.
Direct fulfillment flips the timing.
Because you pay per order, after the customer has already paid you, your inventory turns over far faster, often in under 30 days. Izzy describes the difference as going from a "stop start, stop start" cycle to one that runs "go go go, like this washing machine": make it, sell it, get paid, reinvest.
That same brand goes from $330K tied up to about $75K:
That's roughly a quarter of a million dollars you can put into product or ads instead of leaving it on a shelf.
In fact, Portless says some of its brands push the cycle negative.
Matias Belete, who runs the streetwear brand Foreign Resource, said: "I can get paid before I'm paying for shipping." Even Devin, who argues for keeping inventory at home, gives it that much: "If it's working capital or cashflow, that's a real conversation."
🚨 The cheaper shipping myth: It's easy to assume the whole appeal here is lower shipping costs, but that isn't where the advantage comes from. On a per-order basis, shipping direct from China runs about the same as fulfilling from a US warehouse.
Here's an example.
Take a $30 item that weighs about a pound. Duty is the same either way now that the US has ended de minimis, so set it aside and compare the rest:
These figures combine Portless's published rates with public air and domestic freight benchmarks. Your numbers will vary by product and 3PL.
As you can see, the totals land in the same range. Direct isn't cheaper to ship. What it changes is when you pay: as each order sells, instead of months before.
Side note: other ways to free up cash
Cash flow is the main draw, but it isn't unique to direct fulfillment. A few other approaches can free up cash too.
Bonded warehouse (or a foreign-trade zone). A bonded warehouse lets you import in bulk by ocean but hold off on duty until the goods actually sell, so it captures much of the cash benefit while keeping your stock close to the customer.
In practice:
You bring goods in by sea, the cheap way, instead of flying every order.
You pay duty only as items sell, not all at once when the container lands, which recovers a big chunk of the cash benefit.
Your stock is already in the country, so delivery stays fast and returns stay simple.
Anything that doesn't sell can usually be re-exported without ever paying duty.
It isn't perfect. A bonded warehouse defers duty rather than reducing it, the foreign-trade-zone version has gotten less generous under recent tariff rules, and with so many importers now chasing duty deferral, bonded space itself is in short supply, harder to land and pricier than it used to be.
Even so, for many brands it's the sweet spot, and it's the route Devin's team at ShipMonk leans on. As he puts it, a bonded setup gets you "most of the duty-deferral benefit without the customer-experience penalty."
A few other levers do similar work:
A hybrid setup. Bestsellers sit in a US warehouse for speed; the long tail and new launches ship direct. Fast delivery on the volume that matters, cash efficiency on the rest.
Better factory terms. Net-30, 60, or 90 with the supplier means paying after the goods land, or even after they sell.
Smaller, more frequent runs. Ordering less, more often, holds fewer days of inventory at a time.
Inventory financing. Borrowing against stock or future sales keeps the inventory off your own cash. It costs a fee, but it buys time.
Beyond cash: two more reasons brands do it
Cash flow is the main draw, but it isn't the only one. Two other benefits come up often:
Test products without a big upfront bet
When your inventory sits near the factory and you air orders in small batches, you don't have to commit to a large production run to find out whether something sells. You make a little, air the orders as they come in, and watch what works. The winners you scale. The flops cost you a small run instead of a warehouse full of stock you have to discount your way out of.
For a brand that launches often, that's a real edge: you learn what sells faster, and you're not sitting on dead inventory while you wait to find out.
Reaching new countries gets easier
Normally, selling into a new country means putting inventory there first: a local warehouse, a demand forecast before you have any sales, and stock tied up in a market you're not sure about yet. Here the inventory stays in one place near the factory, and you air orders wherever they come from. Opening a new country becomes closer to switching it on than building a supply chain for it.
Portless says its customers went from shipping to fewer than two countries on average to around seven within nine months.
What you give up: control
The flip side of all that freed-up cash (and other benefits) is control.
When your inventory sits an ocean away and every order flies in on its own, you take on a set of problems a US warehouse would have handled for you.
Returns get harder. Sending a return back to China costs more than the product is worth, so the default move is to refund the customer and write the unit off. Do that at scale and two things happen: you train customers to treat the product as disposable, and you lose the exchange a domestic brand would have captured (Devin puts that at 30 to 40%). As he says, the cost "shows up six months later in repeat rate, not in landed cost." The way around it is to run returns locally, a domestic address that receives, inspects, and restocks. That's the piece Izzy says Portless already handles: "you can return it locally," and from there it's either reverse logistics or a forward-ship to the next customer. It's cheap per unit, about $2 a return, but it's a real operation that has to exist.
Delivery is slower. Five to eight days, against one or two from a US warehouse. Fine for most products, but a problem if your customers expect two-day shipping, or if you sell gifts and seasonal goods where a missed window kills the sale. Setting the expectation at checkout, or framing a launch as a pre-order, takes some of the sting out. Devin makes the same point: done well, a pre-order with a clear ship window "turns delivery time from a complaint into a feature." It still doesn't move the clock.
You react slower. Your stock is close to the factory and far from the customer. That helps when there's a defect to fix or a winner to reorder, and it hurts when something takes off. Devin has "watched brands miss seven-figure demand windows because the next container was still on the water."
It may not suit your brand. "Ships from China" is an asset for a value brand built on cutting out the middleman, and may be a liability for a premium one, where the longer wait and a customs-taped box can undercut the image you've paid to build.
The soft costs add up. Slower, farther-flung orders throw off the things that quietly eat margin: chargebacks from delays, more customer-service tickets, and acquisition spend wasted on customers who don't come back after one slow shipment.
Every parcel is now an import. Each order is a customs entry, with an HTS code, an importer of record, and a customs bond behind it. A capable 3PL handles the mechanics, but the responsibility is still yours, and a slip is how a customer ends up with a surprise duty bill at the door.
If you run a hybrid, your inventory splits in two. Keeping some stock domestic alongside the overseas node means inventory in two places and routing rules to decide which orders ship from where, coordination a single warehouse never needed.
If you use a specialist like Portless, most of this gets handled for you: customs, returns, the overseas node, and last mile. What's left is a brand-level call: the delivery promise you make to customers, and whether the model fits how you sell.
How Quince grew on the model
To see how all of this plays out at scale, look at Quince, the brand that built an entire business on direct fulfillment.
Quince launched in 2020 with one product and a contrarian pitch.
The product was a $50 cashmere sweater that looked like the ones selling for $200. The pitch was that most of what you pay for premium basics is markup, and that shipping straight from the factory to the customer strips it out.
What makes it work is a deal Quince struck with its factories. Instead of buying inventory and warehousing it, Quince convinced them to own the stock and ship one unit at a time off the factory floor. Founder Sid Gupta has called that the hard part to pull off.
Quince does two things:
It keeps the inventory risk on the factory, not on the brand.
It lets Quince test a product with a small run, watch what sells, and scale only the winners, with production cycles measured in weeks rather than months. The company says it sells through nearly everything it makes, so there's little dead stock to mark down.
Its fulfillment runs the way we described earlier:
An order comes in,
It's picked at the factory and consolidated at a hub in Hong Kong.
From there it flies into the US.
Quince prepays the duty, so the customer never gets a surprise bill.
A domestic carrier handles the last mile.
The result is the cash engine from earlier: goods made, sold, paid for, and reordered, with little capital frozen in stock along the way.
It worked at a scale few expected. By 2026 Quince was past a billion dollars in revenue and raised at a $10.1B valuation, all on that factory-direct backbone.
But Quince didn't stay purely cross-border.
As it grew, and especially as it pushed into returns-heavy categories like apparel, it started building US infrastructure: a distribution facility in New Jersey, some inventory held stateside, returns handled at home. That wasn't a retreat from direct fulfillment so much as the model maturing, shipping direct where it shines and keeping stock close to home where it has to.
That arc is the lesson.
The takeaway: direct fulfillment is a strong engine to grow on, and you can rent it from a 3PL instead of building it. Start light, scale the winners, and add the domestic pieces, returns first, as volume climbs. The economics only improve with size, since the best air rates and factory terms come with volume.
How to tell if it’s for you
Start here: check whether your product clears these tests:
1. Does your product weigh ~3.5 pounds or under?
Air freight bills by the pound, so a light product ships cheaply and a heavy one doesn't. Chris de Jong at Portless puts the practical ceiling around 3.5 pounds. If your product weighs more than that, this probably isn't the model for you, and you're better off shipping by ocean and holding stock near your customers.
2. Is your margin above ~40%?
Fulfilling an order this way runs roughly $8 to $9 all-in, which on a $30 order is already close to a third of the price. Add the returns you sometimes write off instead of restocking, and a thin margin disappears fast. Above ~40% you have room to cover all of it and still make money. Below it, you probably don't.
3. Is your average order above ~$25?
There's a fixed cost to flying and clearing every parcel, no matter what's in it. On a cheap order it's a big chunk of the sale. The line sits around $25. The exception is ultra-cheap, rarely-returned commodities like phone cases and cables, where volume carries it anyway.
Clear all three and the direct fulfillment model can work for you.
4. Then look at your return rate
The three tests tell you whether the model can work. Your return rate tells you how hard it'll be to run. The question is just how often your product comes back.
If it rarely does, think jewelry, accessories, small electronics, you're in the sweet spot. The order flies out, lands in a few days, the customer keeps it, and your cash is freed up far sooner than usual. Not much to overthink.
If it comes back often, as most apparel, footwear, and beauty does, the model can still work, but only if you set up returns before you scale. This is where the experts split. Izzy says returns are a solved problem: a good provider runs a local returns node, and that's that. Devin says returns are exactly where brands get burned, because most never build that node and end up eating the loss. They're both right. The model lives or dies on local returns, so that's what you build first.
A few more things worth thinking about
None of these make or break the decision, but the more of them are true, the more the model tips your way:
Your margins are north of 60%, not just past the 40% floor.
Your average order is $75 or more.
You sell, or want to sell, beyond the US.
Cash, not demand, is what's capping your growth right now.
Read it as a dial, not a checklist. A brand that ticks all four is a very different case from one that scrapes through the gates and ticks none.
Beyond the numbers
Everything so far is about your product. The rest is about your brand and your customers, and it matters just as much.
Are your buyers okay waiting five to eight days instead of two? Does a longer shipping window fit the way your brand presents itself, or work against it? And if an order goes wrong, can you fix it without dragging the customer through a hassle? None of this shows up in a weight or a margin, but it's often what decides whether the model actually feels right for your business.
The bottom line
Direct fulfillment is a financing decision more than a logistics one.
You trade a little control, slightly slower delivery and returns to sort out, for a lot of cash freed from inventory. For a light, well-margined product with returns you can handle, that's a great trade. For the wrong one, it isn't.
And you don't have to build any of it yourself. Providers like Portless run the whole model, fulfillment, air freight, and returns, so getting started usually means picking the right partner, not standing up a warehouse.
Which leaves one real question: not whether the model works, but whether it works for you, your product, your brand, your customer promise.
