Where to buy products to sell on Amazon depends almost entirely on how much capital you have, not on which channel is best. With under $1,000 you are doing retail or online arbitrage. Between $1,000 and $5,000, wholesale. Above roughly $5,000, private label from a manufacturer becomes viable. Anyone recommending a single channel regardless of your budget is selling you something.
Here is what each channel actually costs to enter, what margin it returns, and where it stops working.
The six channels, compared honestly
| Channel | Capital to start | Typical gross margin | The real constraint |
|---|---|---|---|
| Retail arbitrage | $200 and up | 25 to 40% | Your own time, and it does not scale |
| Online arbitrage | $500 and up | 15 to 30% | Gating, and price changes between order and delivery |
| Wholesale | $1,500 and up | 15 to 25% | Getting an account approved, and brand restrictions |
| Private label | $5,000 and up | 35 to 60% | MOQ risk and 60 to 120 days of dead capital |
| Liquidation and pallets | $500 and up | Wildly variable | Condition unknown until it arrives |
| Dropshipping | Under $500 | 10 to 20% | Amazon’s policy, which most models violate |
Those margin figures are before advertising and before returns. In our experience sellers plan against gross margin and get surprised by net, so take five to fifteen points off whichever row you are looking at.
Retail arbitrage: the cheapest way to learn
Buying clearance stock from physical stores and reselling it on Amazon. Target, Walmart, Home Depot, TJ Maxx, grocery clearance aisles.
I’d genuinely recommend this as a first channel, and not because it is a good business. It is not. It is the fastest way to learn how Amazon works with almost no capital at risk, and that education is worth the low margin. You will learn gating, fees, the Buy Box, how returns are handled, and how it feels to have inventory sitting there.
Then you will hit the ceiling, because retail arbitrage has one: your own legs. Every unit requires you physically standing somewhere. There is no version of this that scales, and I’ve found sellers who stay in it past the first year usually stay stuck.
Use the Amazon Seller app to scan barcodes in-store. It shows you the current price, the fees, and whether you are allowed to sell the brand at all. Scan before you buy, every time.
Online arbitrage: retail arbitrage that scales a little
Same model, sourced from other websites instead of stores. Better because you can do it at night from a laptop. Worse because everybody else can too.
The specific risk here is timing. You find a product at $12 that sells for $30 on Amazon, you order 40, and by the time they arrive three other sellers have found the same deal and the price is $19. That happens constantly, and it is why I’d never commit more than a few hundred dollars to a single online arbitrage buy.
Watch for gating especially hard on this channel. Many of the best-looking margins are on brands you cannot list without approval, and finding out after the stock arrives is an expensive lesson. Our guide to getting ungated in restricted categories covers what approval takes.
Wholesale: the channel most sellers should be in
Buying genuine stock in quantity, directly from a brand or an authorized distributor, at a real trade price.
This is the channel I think is most underrated. You are selling products that already have demand, reviews, and ranking, so you skip the hardest part of private label entirely. The margins are thinner but the risk is far lower, because you are not betting on whether anybody wants the thing.
The barrier is not money. It is getting approved. Brands and distributors want a business entity, a resale certificate, sometimes a physical address that is not your house, and a reason to believe you will not damage their pricing. Expect to be turned down often, and expect a proper approach to work better than a form fill.
What a proper approach looks like: contact the brand rather than a distributor, tell them specifically how you will present their product, and ask about their Amazon policy before asking for a price list. Many brands restrict Amazon sales entirely, and knowing that in the first email saves you weeks.
Private label: highest margin, highest risk
Having a manufacturer produce your own branded version, usually from Alibaba or a sourcing agent, then building the listing from nothing.
The margins are real and the risk is real in the same proportion. When I model this for a client the number that matters is not margin, it is days of dead capital. Money leaves your account when you place the order. It comes back when units sell. Between those two events sits 30 to 60 days of production, 30 to 45 days of freight, and however long your launch takes to gain traction. Ninety to a hundred and fifty days is normal.
That is the whole reason private label needs $5,000 rather than $1,000. Not the unit cost. The gap.
Order samples from at least three suppliers before committing. I’d treat any supplier who will not send a sample as a supplier who does not exist. And negotiate the minimum order quantity, because the first number quoted is rarely the real floor. Our guide on negotiating with suppliers and moving stock from Alibaba to Amazon FBA covers that process in detail.
Liquidation and pallets: a gamble with a spreadsheet attached
Buying returned, overstock, or shelf-pull inventory by the pallet from liquidation marketplaces or directly from retailers.
Sometimes this is extraordinary and sometimes the pallet is unsellable, from the same source weeks apart. That variance is the channel rather than a bad run, and I’d only enter it with money you can genuinely write off.
If you try it: buy manifested pallets rather than unmanifested ones, at least until you know what you are doing. A manifest lists what is supposed to be inside, which at least lets you price the gamble. Assume 20 to 30% is unsellable. And have a plan for the unsellable portion before it arrives, because storage costs money and a garage full of broken returns is how people quit.
The other thing I’d insist on is checking the source of the returns. Pallets from a single retailer in a single category behave far more predictably than mixed lots, and mixed lots are what get advertised at the attractive prices. In my experience the cheap pallet is cheap for a reason that becomes obvious about an hour into unpacking it.
Dropshipping: mostly a policy problem
Listing a product you do not hold and having a third party ship it to the customer.
Amazon permits dropshipping only under conditions that most dropshipping models break. You must be the seller of record, you must appear as the seller on all packing slips and invoices, and you must handle returns. Shipping directly from a retailer like Walmart to an Amazon customer violates this, because the customer gets a Walmart packing slip.
I’ve found this is the single most misunderstood policy on the platform, largely because a lot of course content sells the version that gets accounts suspended. Read Amazon’s dropshipping policy in Seller Central before you build anything on it.
Legitimate dropshipping, where a wholesaler ships blind under your brand, does work. It is much closer to wholesale than to what is usually marketed as dropshipping.
How I’d actually choose
Answer three questions honestly, and I’d write the answers down rather than think them.
How much can you lose without it mattering? Not how much you have. How much you can write off. That number sets the channel more than anything else.
How much time per week, reliably? Arbitrage is time-heavy and capital-light. Private label is the reverse. Wholesale sits in the middle and is the most forgiving of an inconsistent schedule.
Do you want a business or a brand? Arbitrage and wholesale are businesses that make money and cannot be sold for much. Private label builds an asset with a resale value. Both are legitimate. They are not the same goal, and in my experience choosing between them early prevents a lot of wasted effort.
If you are still deciding what to sell rather than where to buy it, start with how to find low competition products instead, because sourcing a product nobody wants is an expensive way to practice logistics.
FAQ
Where do Amazon sellers buy their products?
Six channels account for almost all of it: clearance stock from physical retail, other websites for online arbitrage, brands and distributors for wholesale, overseas manufacturers for private label, liquidation marketplaces for pallets, and blind-shipping wholesalers for compliant dropshipping. Which one suits you is set mainly by available capital.
What is the cheapest way to start sourcing products for Amazon?
Retail arbitrage, at roughly $200. Clearance stock from local stores, scanned with the Amazon Seller app to check price, fees and whether the brand is gated. Margins run 25 to 40% but the model does not scale, because every unit needs you physically in a store.
How much money do I need for private label on Amazon?
Around $5,000 as a realistic floor, and the reason is timing rather than unit cost. Cash leaves when you place the order and returns when units sell, with 90 to 150 days of production, freight and launch in between. That gap is what the capital covers.
Can I dropship on Amazon?
Only if you are the seller of record, you appear as the seller on every packing slip and invoice, and you handle returns. Shipping directly from a retailer to an Amazon customer breaks this, because the customer receives that retailer’s paperwork. Blind-shipping wholesale arrangements are compliant and behave more like wholesale.
Is wholesale better than private label for Amazon?
Lower risk and lower margin. Wholesale sells products that already have demand, reviews and ranking, so you skip the hardest part of a private label launch. Private label earns more per unit and builds a sellable asset, but ties up capital for months on an unproven product.
Why did my arbitrage margin disappear after I bought the stock?
Because other sellers found the same deal. Prices on arbitrage products move fast once a source becomes visible, and the gap between ordering and listing is usually enough for several sellers to arrive. Keeping individual buys small is the only real defense.
Last updated: September 10, 2026. Amazon fees, gating rules and the dropshipping policy change, so verify current terms in Seller Central and against Amazon’s own seller documentation before committing capital. Margin figures are typical ranges from accounts we work on, not guarantees. ZonHack is an Amazon Ads verified partner and an Amazon SPN Verified Partner.
