how to sell on amazon without inventory

What "selling without inventory" really means

On Amazon, there are really two versions of this. One is inventory-free in the truest sense: you never own the product at any point. Dropshipping and print-on-demand fall here, because a third party manufactures and ships only after a sale happens.

The other is warehouse-free, not inventory-free: with FBA, you still own the goods and they still count as your inventory for accounting and tax purposes. You just don't store or pack them yourself. Amazon does. Sellers often conflate the two, and that confusion causes real accounting errors later, especially around cost of goods sold and reimbursements.

Why does the distinction matter so much? Because your balance sheet and your tax return treat the two cases completely differently. In a true no-inventory model, product cost hits your books as an expense the moment the supplier invoices you for a shipped order. In the FBA model, the money you spend on stock sits on your balance sheet as an inventory asset until each unit sells, and only then does it move to cost of goods sold. Get that wrong and your profit numbers swing wildly month to month for no real reason, which makes it impossible to tell whether the business is actually working.

That split has a name: periodic vs perpetual inventory accounting. Periodic means you count what's left at the end of a period and back into cost of goods sold from the difference. Perpetual means each unit's cost moves out of inventory and into COGS at the moment it sells. Dropshipping and print-on-demand are effectively perpetual by default, since there's nothing sitting on a shelf. FBA sellers need a perpetual system too, because units sell every day across dozens of SKUs and a quarterly count can't tell you which listing is actually making money.

Four ways to do it, and what each one costs you operationally

Pick based on margin tolerance and how much control you want over quality and fulfillment speed.

• •Dropshipping through a manufacturer or verified wholesale supplier: you list, customer orders, supplier ships direct. Requires airtight supplier reliability since Amazon holds you responsible for delivery and condition, not the supplier.
• •Print-on-demand / Merch on Demand: a vendor prints and ships apparel, mugs, or books only after a sale. Zero upfront stock risk, but margins are thin and design differentiation matters more than sourcing.
• •Kindle Direct Publishing (KDP): digital or print-on-demand books with no physical stock held by you at all.
• •Wholesale or private-label selling fulfilled by Amazon (FBA): you buy stock from a manufacturer, ship it to Amazon, and Amazon warehouses, picks, packs, and ships it. This is the most scalable model but the one people most often mislabel as "no inventory." You own it, Amazon just stores it. Many sellers never handle the boxes here either, because a prep center for Amazon FBA can receive the supplier's shipment, inspect it, label and poly-bag units, then forward everything to the fulfillment center for you.

How the money actually moves in each model

The cash flow patterns differ more than most guides admit, and the differences decide which model fits your bank balance. In dropshipping, you collect the customer's payment through Amazon, then pay the supplier per order. Amazon settles your sales to your bank roughly every two weeks, but your supplier usually wants payment immediately, so you're floating the supplier cost out of your own pocket for days or weeks on every single order. Thin margins plus that timing gap is how dropshippers with "profitable" spreadsheets run out of cash.

Print-on-demand flips less cash but also returns less. The vendor charges you a base production cost per unit sold, Amazon takes its referral fee, and whatever remains is yours. There's no order-minimum risk, which is why POD is a popular first model, but there's also no volume discount lever to pull as you grow. Your margin at unit 1 looks a lot like your margin at unit 10,000.

FBA wholesale front-loads everything. You pay a manufacturer for hundreds or thousands of units before a single one sells, pay freight to get them to Amazon's warehouse, then wait for sell-through. The reward is control and scale; the cost is that a bad buying decision ties up cash for months. Say you spend a hypothetical $8,000 on 1,000 units at $8 landed cost each. If they sell for $20 with roughly $8 in combined Amazon fees per unit (check the current fee schedule for real numbers, and run your own SKUs through an FBA calculator before committing), you'd gross about $4 per unit. That's $4,000 of profit potential, but only if all 1,000 sell before storage fees and price erosion eat into it. The same math with 30% of units stranded unsold looks very different.

The accounting catch nobody mentions

Even in a true no-inventory model, you need a reliable per-unit cost. Tax reporting requires cost of goods sold, and COGS math doesn't change: Beginning Inventory + Purchases − Ending Inventory = COGS. If you're dropshipping, your "purchase" is the supplier invoice per order. Track it per SKU, not just as a lump sum, or your margin reporting becomes guesswork.

If you go the FBA wholesale route instead, the stakes get higher. Amazon's warehouses lose, damage, and misplace units, and getting reimbursed correctly depends entirely on the cost figure on file. Since 2025, Amazon reimburses lost or damaged FBA inventory based on the seller's manufacturing or sourcing cost, not the retail price, and Amazon will use its own estimate of that cost unless you've supplied your actual number. Sellers who never uploaded accurate landed costs routinely get reimbursed at a fraction of what the unit actually cost them.

That single policy change is why sellers who think they've sidestepped inventory accounting by using FBA still need clean cost records. The FBA inventory reimbursement policy makes your own cost documentation the deciding factor in how much cash you get back when Amazon's warehouse makes a mistake, and warehouse mistakes are frequent enough that this isn't a rare edge case. Some sellers use amazon reimbursement software or a dedicated fba reimbursement service to audit settlement reports and file these claims; others do it manually. Either way, the claim is only as strong as the cost data behind it.

Common mistakes that sink no-inventory sellers

The failure patterns are predictable, which means they're avoidable. Here are the ones that show up over and over in sellers' books.

Treating Amazon's deposit as revenue is the big one. The bank deposit you receive is sales minus referral fees, FBA fees, refunds, advertising charges, and a dozen smaller adjustments. If you record the deposit as your revenue, you understate both sales and expenses at the same time, and your gross margin becomes fiction. Every settlement needs to be broken out line by line, which is tedious by hand and exactly the kind of thing automated accounting software exists to do.

The second mistake is ignoring returns in margin math. A dropshipped or POD return often can't go back to the supplier, so the full product cost is a write-off, not just the fees. A model that looks profitable at a 2% return rate can be underwater at 8%. Amazon also grants some low-value claims as a refund without return, where the buyer keeps the item and the refund still comes out of your proceeds. Nothing comes back into stock, so the loss never appears as a returned unit, only as a hole in the margin you thought you had. Build a returns assumption into your per-SKU margin before you scale a listing.

Third: mixing personal and business spending, then trying to reconstruct COGS at tax time from a shoebox of supplier invoices. Amazon bookkeeping done retroactively is slower, more expensive, and less accurate than logging costs as they happen. A separate bank account and a cost log per SKU on day one costs you an hour of setup and saves you a brutal January.

Getting your books ready before you scale

Whichever model you pick, set up cost tracking before your first sale, not after your first tax season. For dropshipping and POD, log the supplier cost per unit at the SKU level. For FBA wholesale, log landed cost (unit price, freight, duties, prep) and keep it current in Seller Central so Amazon's reimbursement engine has your real number instead of its own estimate.

This is exactly the gap Amazon accounting software is built to close: reconciling settlement data, COGS, and reimbursement shortfalls automatically instead of leaving you to rebuild it from spreadsheets after the fact. Don't wait until reconciliation season to discover your cost basis was never recorded.

What should you actually look for in tooling? For a low-volume dropship or POD side business, honestly, a spreadsheet plus QuickBooks or Xero can carry you for a while, and paying for more is premature. The tipping point comes when settlements get too dense to break out by hand, usually somewhere in the low hundreds of orders a month. At that point, accounting software for amazon sellers should do three specific things: split each settlement into revenue, fees, and refunds automatically; keep a perpetual per-SKU cost record so COGS posts as units sell; and flag reimbursement discrepancies where Amazon lost stock and paid you less than your documented cost. Amazon quickbooks integration or an amazon xero integration handles the ledger side; the seller-specific layer on top (BeanHawk is one option, A2X and Link My Books are others worth comparing) handles the settlement parsing and cost tracking. If you're running FBA at any real volume, add ecommerce inventory management software or at minimum a disciplined landed-cost sheet, because your reimbursement claims and your tax return both depend on the same cost numbers agreeing with each other.

One caution on the inventory side of that stack. Most small business inventory software is designed around a warehouse you control: bins, pick lists, purchase orders, and a receiving desk. Small business warehouse inventory software solves a problem you don't have when Amazon or a print vendor holds the goods. QuickBooks Online tracks quantities and average cost on its higher plans, and the QuickBooks Online app store lists inventory apps (SOS Inventory and Katana come up often) that extend it further. Cheap inventory management software, or even a clean sheet, is genuinely fine for a POD business whose only real cost is the vendor's per-unit charge. The test for an Amazon seller isn't how many invoice and sales features a tool ships with, it's whether it holds a per-SKU landed cost and moves that cost to COGS when the unit sells. A simple system fed with accurate numbers beats an expensive one fed with guesses.