Amazon Inbound Shipping: STA, Carriers, and AWD in 2026

There’s a number on the Send to Amazon screen that you’re supposed to ignore. Amazon calls it the estimated shipping cost, it sits right under the placement options, and if you’re shipping DDP with your own freight forwarder it has almost nothing to do with what you’ll actually pay. We ignore it on every client shipment. Amazon inbound shipping is where a lot of first launches quietly hand over a few hundred dollars nobody budgeted for, usually to a placement fee that was never compared against a freight quote. Three decisions control that cost, and only one of them is obvious.

01The three names, and the one that doesn’t exist

Sellers throw three acronyms around when they talk about getting boxes into Amazon. Two of them are real programs. One of them is a ghost, and it’s worth clearing up before you go looking for a setting that isn’t there.

STA is Send to Amazon. It’s the workflow inside Seller Central that turns “I have 500 units at a factory in Ningbo” into a shipment plan, a set of box labels, and a ship-to address. It replaced the older Manage FBA Shipments screen. STA isn’t a service you opt into and it isn’t a price. Every FBA shipment you ever create goes through it, so calling it an option is a bit like calling the checkout page an option.

AWD is Amazon Warehousing and Distribution. That one is a real, separate service. Think of it as your own bulk storage warehouse that Amazon happens to run. Storage there is cheaper per cubic foot than a fulfillment center, and the critical thing to understand is that AWD does not ship to your customers. It feeds FBA. Nothing leaves an AWD building and lands on a doorstep.

The third one, PBA, doesn’t exist as an Amazon program. I’ve seen it in seller forums and in a few blog posts, usually as shorthand for “partnered by Amazon,” and people use it when they mean Amazon’s partnered carrier option. That option is real. The acronym isn’t. If you go hunting for a PBA dashboard in Seller Central you’ll waste an afternoon.

So the three real decisions at the inbound step are:

1. How you split the shipment. Amazon offers you a few placement options with different fees attached. This is where most of the surprise cost lives.
2. Who carries the boxes. Amazon’s partnered carrier, or your own freight forwarder as a non-partnered carrier.
3. Where the stock lands. Straight into FBA, into AWD for bulk, or a split of both.

Everything else on those screens is data entry.

02Where the money actually goes at the inbound step

Amazon started charging an inbound placement service fee in March 2024, and it changed the arithmetic of a first shipment for anyone sending a single consolidated pallet. The logic is simple enough: Amazon wants your inventory spread across several fulfillment centers so it’s close to buyers. If you refuse to split it, Amazon does the spreading itself and bills you for the internal transfer.

The fee is per unit, and it scales with two things: how big and heavy the item is, and how few destinations you agree to. Amazon updated the rate structure on January 15, 2026, adding new weight bands and a Small Bulky tier. Roughly speaking, a standard size unit sent as a minimal split costs somewhere in the region of 30 to 40 cents per unit at the light end, heavier standard units climb past 70 cents, and extra large items can reach a couple of dollars each. Rates move, and third party fee trackers disagree with each other more often than you’d like, so pull the live rate card from your own Seller Central account before you build a budget on any number you read online, including this one.

The option that costs nothing is the one where Amazon decides. If you accept the Amazon optimized split, usually shown as the largest number of destinations, the placement fee is $0.00. You trade that saving for a bigger freight bill, because your forwarder now has to deliver to several addresses instead of one.

That’s the whole tension, and it’s a real arithmetic problem rather than a rule. On a 500 unit order at 20 units per box, you’re moving 25 cartons. Take a minimal split at 40 cents a unit and you’ve added $200 to the shipment before a single box moves. If your forwarder quotes less than $200 extra to hit five addresses instead of one, the free option wins. If they quote $350 extra, it doesn’t. You cannot answer that from the Amazon screen alone, which is exactly why so many sellers get it wrong.

Two more costs sit quietly in the background. FBA storage is billed monthly per cubic foot and jumps hard in Q4, which we covered in detail in our Amazon FBA storage fees breakdown. And prep, since January 2026, is entirely on you, which changes who touches the boxes before they ever reach this stage.

One more thing worth knowing before you get to volume: AWD shipments don’t carry a separate placement fee at all. Placement is folded into what you’re already paying for the storage service. That barely matters on a first 500 unit order, because you probably want all of it live in FBA anyway. It matters a lot on a 3,000 unit reorder where two thirds of the stock is going to sit somewhere for four months regardless. At that size the placement question stops being a line item and turns into a routing decision.

03The Send to Amazon workflow, seven steps

Here’s the actual click path we walk clients through. It looks longer written down than it feels the second time you do it.

1. Pick the SKU. Seller Central, three line menu, Inventory, Manage All Inventory. Tick the box next to the product, then choose Send or replenish inventory from the menu that appears at the bottom.
2. Confirm the ship from address. This is where the inventory starts its journey, so it’s usually your factory in China or your freight forwarder’s warehouse, not your home address.
3. Build a packing template. The field that matters most here is units per box. If your supplier packed 20 units into every carton, enter 20. Then the outer box dimensions and the packed box weight, measured on the carton, not the product. If your FNSKU labels are already applied at the factory, select no prep needed.
4. Enter the quantity. You enter boxes, not units. Amazon multiplies boxes by your units per box to get the total, so 25 boxes at 20 units each shows as 500 units. Then Ready to send, then Confirm and continue.
5. Choose your carrier and placement option. Two carrier choices, several placement options, and the estimated shipping cost you’re about to ignore. More on this in the next section.
6. Accept charges and confirm. You’ll pick a shipping mode, air or ocean, and lock the plan.
7. Print the box labels. Choose thermal 4x6 inch format, which is what nearly every factory and forwarder can print. Download the PDFs and email them to whoever is packing your boxes.

That last step has a trap in it. If Amazon split your shipment into three destinations, you get three separate label files. Your factory now needs to be told, clearly and in writing, which label file goes on which boxes. Get that wrong and cartons land at the wrong building, which turns into a reconciliation case rather than a receiving event.

Name the packing template properly while you’re in there, because you’ll reuse it. If every reorder of the same SKU ships 20 units to a carton at the same dimensions, you build that template once and select it on every future shipment. Sellers who call theirs “Template 1” end up with six of them and no idea which is which. Name it after the SKU and the units per box. The prep category field inside the template also carries more weight since January 2026: if your factory applies the FNSKU labels, select no prep needed, and confirm that’s actually true before you tick it.

04Partnered carrier or your own forwarder

Step 5 asks you to pick between an Amazon partnered carrier and a non-partnered carrier. On client shipments coming out of China, we pick non-partnered almost every time.

Partnered means Amazon arranges the pickup through a carrier it has a deal with, typically UPS for small parcel or a partnered LTL line for pallets. On paper it’s convenient. In practice, for an international shipment, you’re still the one handling the customs declaration and the duties, and the landed number has consistently come out higher than what we get through our own forwarder. It’s a reasonable option for domestic restocks from a US warehouse. It’s rarely the right one for a first container out of Asia.

Non-partnered means you tell Amazon “somebody else is bringing these,” and your freight forwarder does the rest. On our client launches we ship DDP, which means the quote we hand the client already contains the ocean or air freight, customs clearance, duties, and delivery to the fulfillment center. One number, no port surprises. The 500 unit vegetable chopper order we’ve written about elsewhere landed at about $4,500 DDP, roughly $9 a unit, and that figure already included everything up to the Amazon door.

If you’re collecting DDP quotes for the first time, ask for the full picture rather than a per kilo rate. A usable quote names the freight mode, the pickup point at the factory, whether customs clearance and duties sit inside the number, the delivery address or addresses, and an estimated arrival window at the fulfillment center. A quote that gives you a rate and nothing else tends to grow later, usually at the port, and usually while your listing is sitting at zero inventory.

Which brings us back to the estimated shipping cost on screen. That estimate exists to price Amazon’s partnered carriers. It does not include US customs duties or port fees. If you’re on a DDP quote, those are already paid inside your forwarder’s number, so the Amazon estimate is describing a shipment you aren’t taking. Ignore it and move on.

The other field people misread on that page is the delivery window. It’s not your ship date. It’s the estimated date range the boxes arrive at Amazon. Ask your forwarder for their estimated arrival date at the fulfillment center and use that, because guessing here creates a shipment that shows as late for reasons nobody can trace.

Now the cross check, which is the single highest value habit at this step. Before you accept any placement option, send your forwarder both scenarios and ask for two quotes: what does it cost to deliver to these five addresses, and what does it cost to deliver to this one address. Then add the Amazon placement fee back onto the one address quote. Compare the two totals. Pick the smaller one. That’s it. It takes one message to your forwarder and it’s the difference between a $0 placement fee that quietly costs you $350 in extra trucking, and a $200 fee that was actually the cheaper path. We do this on every client shipment, and it goes both ways often enough that we’ve stopped assuming an answer.

If you’re still choosing a freight method in the first place, that decision sits upstream of all of this and we broke it down in our guide to shipping from China to Amazon.

05Five mistakes we see on first shipments

Across 47 client onboardings since 2022, the same handful of inbound errors keep showing up. None of them are complicated. All of them cost either money or weeks.

Creating the shipment plan too early. A shipment plan is valid for 90 days from creation, then it closes for inactivity. Sellers create the plan the day they place the production order, feel organized, then watch it expire while the factory is still running. Create the plan when the goods are genuinely close to ready, and get the tracking numbers from your forwarder into the shipment tab once the boxes actually move.

Sending the wrong label file to the factory. Covered above and worth repeating, because it’s the most expensive clerical error at this stage. Three destinations means three label PDFs and three explicit instructions.

Treating Amazon’s estimated shipping cost as your cost. It prices partnered carriers only and excludes duties. Sellers see a low number, feel good, and then get a real invoice that looks nothing like it.

Entering units per box wrong. Amazon multiplies boxes by units per box. Enter 10 when your cartons hold 20, and Amazon expects 250 units while 500 arrive. That’s a receiving discrepancy, and discrepancies get resolved slowly.

Not pulling the ship to address before asking for quotes. Your forwarder cannot quote accurately without knowing which buildings the boxes are going to. Create the plan, open it under Inventory then Shipments, copy the ship to address, and send that to your forwarder. Guessing at a generic Amazon address produces a quote you can’t rely on.

There’s a sixth one that isn’t really an inbound mistake but shows up here: bundles. When you add a glove, a manual, or a carrying bag to differentiate a product, the carton gets bigger and heavier, and both your freight cost and your placement fee move with it. That’s not an argument against bundling, since bundle differentiation is one of the few reliable ways to escape a price race. It’s an argument for costing the bundle at the carton level before you commit to it.

06AWD, FBA, and the Capacity Manager

Once volume grows, you stop deciding only how to ship and start deciding where inventory sleeps. Three options, and they’re priced very differently.

The rule we use is simple. If you’re holding more than three months of supply, that excess belongs in AWD, not FBA. Send a working portion to FBA for live sales and park the rest upstream. AWD storage rates are meaningfully lower than fulfillment center rates, and placement is bundled in rather than billed as a separate line.

Creating an AWD shipment feels almost identical to a normal FBA one. Inventory, then Warehousing and Distribution, then Send to AWD, and you’re back in a Send to Amazon style workflow. Box information, box quantity, Ready to send, Confirm and continue. On the shipping step, if your forwarder or manufacturer is moving the goods, which is nearly always the case, select Seller Managed, set the ship date, and choose Other as the carrier if you’re on your own forwarder. Then accept charges, print the AWD labels, and send those specific labels to whoever is packing.

Two operational details that catch people. First, replenishment from AWD into FBA can run automatically or manually. Automatic is a setting: Amazon moves stock down when FBA runs low, which is the whole point of the service. Manual gives you control if your sales are lumpy. Second, after you create your first AWD shipment, the AWD Manage Inventory page can take two to three days to become visible in your account. Nothing is broken. It just hasn’t provisioned yet.

AWD isn’t free optionality, and the tradeoff is worth saying out loud. Stock sitting upstream is not stock a customer can buy today. Replenishment takes time to move units down into a fulfillment center, so if sales spike, the inventory you were being clever about storing cheaply is the inventory you can’t sell this week. That’s why the split matters more than the service itself. Keep enough in FBA to cover realistic sell-through plus a buffer, and send the rest upstream. The saving only counts if you never go out of stock chasing it.

Capacity Manager is the third lever and the one to use last. Amazon moved from unit based storage limits to volume based limits in cubic feet, and each size tier has its own cap. Check yours under Inventory, then Shipments, then the Capacity Monitor tab. If you’re at 100% and need more space, Capacity Manager runs like an auction: you request a volume, you bid a maximum reservation fee per cubic foot, and Amazon charges the clearing price rather than your bid. Bid $5.00 when the winning price was $3.00 and you pay $3.00.

The catch is in the name of the bet. If you sell through the inventory, performance credits usually offset the reservation fee. If you don’t sell it, you owe the fee anyway. Request 1,000 extra cubic feet at $3.00 and sit on it, and that’s $3,000 of exposure for space you didn’t need. Use Capacity Manager only when the stock has to be Prime-ready right now. For everything else, AWD is the cheaper answer.

07When to hand the inbound step over

Most sellers can run their own first shipment. It’s fiddly rather than hard, and doing it once teaches you more about your own cost structure than any article will. The step that trips people up isn’t the clicking, it’s the cross checking: getting two freight quotes, adding the placement fee back, and making a decision under time pressure while a factory waits on labels.

The point where handing it over starts paying for itself is usually the third or fourth shipment, when you’re running multiple SKUs, splitting between FBA and AWD, and every week of delay is a week of lost rank. That’s the stage our done for you clients are at. We create the shipment plans, run the placement cross check with our own forwarder, and get the labels to the factory so the boxes move on schedule rather than on whoever remembers. Prep sits directly upstream of this and has its own decision tree, which we walked through in our guide to FBA prep centers versus self prep. If you want a look at your specific setup before your next order ships, book a discovery call and we’ll go through it with you.

Frequently asked questions

What is Amazon inbound shipping?

Amazon inbound shipping is everything between your supplier’s loading dock and an Amazon warehouse receiving your boxes. It covers creating the shipment plan in Send to Amazon, choosing a placement option, picking a carrier, printing box labels, and getting the freight delivered. It does not cover Amazon shipping to your customer, which is the fulfillment side.

What’s the difference between STA and AWD?

Send to Amazon is the workflow you use to create any shipment. AWD is a separate bulk storage service that holds inventory upstream and replenishes FBA. You use Send to Amazon to create shipments into both. The plain difference is that FBA stock ships to buyers and AWD stock does not.

Is there an Amazon program called PBA?

No. There’s no Amazon inbound program with that name. Sellers who write PBA are usually referring to Amazon’s partnered carrier option, which appears inside the Send to Amazon workflow when you choose who transports your boxes. The alternative is a non-partnered carrier, which means your own freight forwarder.

How do I avoid Amazon inbound placement fees?

Accept the Amazon optimized split, which sends inventory to the largest number of destinations and carries a $0.00 placement fee. Before you do, get a freight quote for that multi address delivery and compare it against a single address quote plus the placement fee. Whichever total is lower is the right answer, and it changes by shipment. AWD shipments include placement, so they don’t carry a separate fee.

When should I use AWD instead of sending everything to FBA?

Once you’re holding more than roughly three months of supply. Storage in a fulfillment center is the most expensive square footage in Amazon’s network, and Q4 makes it worse. Send the portion you expect to sell in the next couple of months to FBA and park the rest in AWD, with replenishment set to move stock down as FBA runs low.

How long is an Amazon shipment plan valid?

90 days from the day you create it. After that it closes for inactivity and you have to build it again. Create the plan when your production order is genuinely near completion, then add the tracking details from your freight forwarder to the shipment tab once the boxes are on their way.

The bottom line

The inbound step looks like data entry and behaves like a pricing decision. Send to Amazon is just the workflow. The two things that actually move money are which placement option you accept and who carries the boxes, and neither can be answered from the Amazon screen on its own. Get two freight quotes, add the placement fee back onto the single destination option, compare the totals, and pick the smaller one. Ignore the estimated shipping cost if you’re on a DDP quote. Once you’re carrying more than three months of stock, move the excess into AWD instead of paying fulfillment center rates for inventory that isn’t selling yet. We run those checks on every client shipment, and they’re the difference between a launch that lands on schedule and one that pays for the same boxes twice.