Operations: fulfilment, packaging & compliance
The short answer is that you usually should not ship to a retailer from FBA, and the reason has almost nothing to do with cost per unit. It is that a fulfillment center produces parcels addressed to consumers, and a retail distribution center is built to receive pallets of cases. Those are two different logistics products, and one cannot be bent into the other.
There is a narrow band where Amazon's own fulfillment can serve a retail customer, and it is worth knowing exactly where that band ends, because the first order from a boutique or an independent chain often sits inside it. Past that, the answer is a warehouse.
What Multi-Channel Fulfillment can and cannot do
Multi-Channel Fulfillment lets you create an order against your FBA stock and ship it to any address. Mechanically it is the same pick and pack that serves an Amazon order, pointed elsewhere. For a small drop-ship order it genuinely works: a boutique orders eighteen units, you key in the address, cartons arrive in a few days, nobody at either end has to think about freight.
Where it stops working is everywhere a real retail account lives.
Packaging is unbranded by default, and even where it is plain, what arrives is a shipping box holding loose units, not a sealed case with a case label. Nobody at the receiving end can scan a GTIN-14 that does not exist.
There is no case pack. The DC's system expects a fixed number of units per carton, matched to what the purchase order says. Multi-Channel Fulfillment packs to whatever fits the box it chose.
There is no pallet. Orders arrive as parcels on a small-parcel carrier, which means they go to the retailer's parcel door if they have one and get refused if they do not.
There is no routing compliance. A retailer's routing guide names which carriers may be used above which weight, which appointment system to book through, and which documents travel with the shipment. Amazon's outbound network does not participate in any of that.
And the per-unit price is consumer fulfillment pricing, which is fine on a single unit and absurd across four hundred of them. A case of twelve moving as twelve individually handled picks costs a multiple of what the same case costs moving on a pallet. The full picture of what the service is and is not is in Amazon Multi-Channel Fulfillment.
Use it for sample shipments, for a first tiny order from an independent store, and for drop-ship orders a retailer places on behalf of a customer. Do not use it to fill a purchase order with a ship window on it.
Why a DC cannot receive a stack of poly-bagged units
Picture the receiving side for a moment, because it explains every rule that follows.
A receiver has a purchase order on screen, an appointment window, and a dock door. A truck arrives, the pallet comes off, the receiver scans one case label and the system tells them what should be inside and how many. They count cases, not units. They sign for the pallet against a bill of lading. The whole transaction takes minutes.
Now hand that person forty parcels of mixed count with an FNSKU sticker over the barcode and no case label. There is nothing to scan and no way to reconcile against the PO without opening everything. That is not a receiver being difficult — their system has no field for it. The shipment gets refused, or received short, or received and then deducted.
This is also why the packaging work described in Amazon seller retail compliance is not optional polish. It is the interface. The pallet is the message.
Do removal orders solve it?
Partly, and slowly.
A removal order sends your FBA units to an address you nominate, usually your own or your warehouse's. What arrives is your stock, in Amazon's prep, in mixed cartons, weeks later. You then have to strip or cover the FNSKU labels, decide what to do about poly bags, count units into a real case pack, seal and label cases, build a pallet and book freight.
That is a legitimate path when you already wanted the stock out of Amazon, and it is the only path when the units you need to ship are already in fulfillment centers and nowhere else. Just cost it truthfully: removal fees, inbound freight to your warehouse, the labor per unit to repack, and the calendar. If the buyer's ship window is three weeks out, a removal usually will not make it.
The structural fix is to stop needing removals, by splitting the production run so retail-ready cases exist before anything reaches Amazon. That is the whole argument in moving Amazon inventory to retail stores.
Where should the order actually ship from?
| Origin | Fits | Where it breaks |
|---|---|---|
| Amazon Multi-Channel Fulfillment | Samples, one small independent store, drop-ship | No case pack, no case label, no pallet, no routing compliance, consumer pricing |
| Removal then reship | Stock already stuck in FBA, slow movers you want out | Weeks of lead time, per-unit repack labor, condition risk |
| Warehouse direct | Any real purchase order, any chain, any reorder | Needs a second inventory pool and a warehouse that ships business freight |
Read the third row as the destination and the first two as detours. Almost every brand that keeps a retail channel ends up shipping from a warehouse, because it is the only origin that can answer a routing guide.
What a third-party warehouse actually does for you
The word warehouse undersells it. Storage is the cheapest thing in the quote. What you are buying is the set of tasks a DC expects on the other end of the phone.
Receiving your container or your factory shipment and putting the retail-ready cases away as full cases. Picking to a purchase order in case quantities, not eaches. Building a pallet to the retailer's spec — the right tier and pattern, stacked square, no overhang past the pallet edge, wrapped so the load arrives as one unit rather than a collapsed pile. Printing and applying carton labels and the pallet placard. Booking the carrier the routing guide names, or handing the shipment to the carrier the retailer's own traffic team routed. Sending a clean bill of lading and, where required, an electronic shipping notice ahead of the truck.
When you shop for one, the question that separates real candidates is simple: have you shipped to this retailer before, and can you send the carton label format you use for them. A warehouse that already ships into that DC has the routing guide, the appointment login and the label template. One that does not will learn on your order, and the tuition is billed to you as deductions.
Parcel or freight: where the line sits
Small parcel makes sense up to roughly a few hundred pounds and a handful of cartons, when the retailer accepts parcel delivery and the destination is a store rather than a DC. Independent retailers and single-store accounts often prefer it, because they have no dock.
Less-than-truckload freight takes over once you have a pallet. A pallet is one handling unit regardless of whether it holds thirty cases or sixty, which is why the per-unit cost falls off a cliff the moment you palletize. Most retail DCs will only accept freight and will tell you so in the routing guide, often naming the carrier and requiring you to ship collect on their account.
The practical sequence for a first freight shipment is worth memorizing, because it is where new shippers get deductions. Cases labeled and counted. Pallet built to the required height and stretch wrapped with the pallet placard applied on two adjacent sides. Bill of lading raised with the purchase order number on it. Appointment booked through whatever portal the routing guide names. Shipping notice transmitted before the truck leaves, if the retailer requires one. Then the pickup.
Miss the appointment and the truck waits or gets turned away. Miss the purchase order number on the bill of lading and receiving cannot match the load. Miss the shipping notice and you get a deduction with a category name attached to it.
Does any of this change what you charge?
It should. Retail freight and warehouse handling are real costs per case, and they belong in the wholesale price rather than in a note-to-self about fixing it later. Work out the fully loaded cost of a case landed at the DC — product, case packaging, warehouse pick and pack, freight, and a provision for deductions — before you quote a buyer, not after.
If you are still deciding whether a retail channel is worth building this machinery for, the argument and the sequence are in how to sell Amazon products in retail stores. And if the open question is which retailers would plausibly take the product at all, paste the listing into WholesalePilot and the preview shows who would stock it, which is a cheaper way to find out than building the supply chain first.
None of this is harder than Amazon's prep rules were the first time you read them. It is just a different set of rules, written by a different kind of receiver, and it starts from a pallet rather than a parcel.