STRATEGY
Retail Arbitrage on Amazon: How to Decide What to Buy
Retail arbitrage means buying products cheaply in a shop and reselling them on Amazon at a higher price. It is the fastest way into selling on Amazon because you need no supplier, no brand and no minimum order. What it does need is a way to tell, in the aisle, whether the thing in your hand is worth the money. This guide covers how the model works, what to check, and the numbers that actually decide it.
What retail arbitrage actually is
You walk into a shop, find something on clearance, and check what it sells for on Amazon. If the gap between the two prices covers Amazon's fees and leaves a worthwhile return, you buy it, send it to Amazon, and list it on the existing product page.
There is no product to develop, no supplier to negotiate with, no brand to build. You are selling something that already exists on a listing that already has customers. That is why it is the usual starting point for new sellers, and why plenty of experienced sellers still do it alongside private label.
Online arbitrage is the same model without the shop. You buy from another retailer's website instead of a shelf. The decision is identical, and everything in this guide applies to both.
The four questions, in order
Experienced sellers do not start with the profit. They work through four questions in a fixed order, and most products are rejected on the first one within seconds.
How a sourcing decision actually runs
- 1. Can I even sell this?
- Category approval, brand restrictions, hazardous goods, meltable stock, and whether Amazon itself is on the listing. If any of these fire, nothing else matters.
- 2. Does it actually sell?
- Not whether it looks popular. Whether units are genuinely moving, and how fast, judged against the category rather than a raw rank number.
- 3. Does the money work?
- Amazon fees, your cost, and the return that leaves. The key figure is the most you can pay and still hit your target.
- 4. What happens after I buy?
- Whether the price is stable, how many sellers are competing, and whether more are arriving. This is where most money is lost.
AskJeffy runs all four checks from a barcode scan. Point your phone at the label and see the answer in seconds. Try AskJeffy free
Question one: can you sell it?
Amazon restricts who can sell in certain categories and who can sell certain brands. Buying stock you cannot list is the most common expensive mistake a new arbitrage seller makes.
What to check before anything else
- Category approval. Grocery, beauty, health, jewellery, watches and several others usually need approval before you can list.
- Brand restrictions. Many brands limit who may resell their products, regardless of category.
- Hazardous goods. Aerosols, batteries, flammables and pressurised items have restricted fulfilment and extra requirements.
- Meltable stock. Chocolate, candles and similar cannot be sent to Amazon between 16 October and 30 April.
- Amazon on the listing. If Amazon sells the product itself, you are competing with the house.
One important caveat: no third-party tool can see your own account approvals without connecting to your Amazon seller account. Tools can tell you a category is usually restricted or a brand is known to limit resellers, but the final check belongs in Seller Central.
Question two: does it actually sell?
A product with a beautiful margin that nobody buys is worse than no product at all, because your money is tied up in it indefinitely.
Sales rank is the usual measure, and the usual mistake. A rank of 50,000 is fast in Home and Kitchen and nearly dead in Industrial and Scientific, because the categories are wildly different sizes. The raw number tells you almost nothing on its own. What matters is the position within that category: most sellers source within the top one to three percent.
Rank drops are better than estimates. Every time a product's sales rank drops, something sold. Counting those drops gives you a floor, a number you can trust, rather than a model's guess at monthly volume. When a tool shows you both and they disagree wildly, believe the count.
Question three: does the money work?
Amazon takes a referral fee, usually 8 to 15 percent depending on category, and a fulfilment fee if you use FBA. Together they typically take a quarter to a third of the sale price before your cost is even counted.
The number that actually decides a purchase is not the profit. It is the most you can pay for a unit and still hit your target return. If the shelf price is below it, the deal works. If not, you put it back. Everything else is context around that one figure.
Return targets sellers actually use
| Type of product | Minimum ROI | Why |
|---|---|---|
| Low cost, under £5 | 50 to 100% | The handling cost is the same as a £50 item, so the percentage has to be higher to be worth doing |
| Standard | 30% | The usual working baseline |
| Grocery and consumables | 35% | Expiry dates and stricter requirements add risk |
| High ticket, over £75 | 18 to 25% | The absolute profit is large enough to justify a thinner percentage |
| Q4 toys and games | 45 to 70% | Prices collapse after the holidays, so you need a buffer |
Do not forget what a bad calculation costs. Getting the fees wrong by a couple of pounds turns a profitable buy into a loss, and you will not find out until the payout arrives weeks later. Our FBA fees guide covers what Amazon actually charges.
Question four: what happens after you buy?
This is the one new sellers skip and experienced sellers care most about, because it is where the money goes.
A listing that looks profitable today can collapse in weeks. It happens when a good deal gets shared: dozens of sellers spot the same clearance, all buy in, all list at once, and the price falls through the floor as they undercut each other. Sellers call it the jaws of death, after the shape it makes on a chart when the seller count spikes up and the price curves down.
The signature is visible before the collapse finishes. New sellers arriving faster than the listing has ever absorbed them, while the buy box price is already falling. Both halves are needed: sellers arriving against a firm price is ordinary competition, and a price dip with no new sellers is usually a promotion.
This is why price history matters more in arbitrage than in almost any other model. A product priced at £26 today that has sat at £18 for most of the last year is not a £26 product. It is an £18 product having a good week.
Amazon on the listing is three situations, not one
Most guides tell you to avoid any listing where Amazon is a seller. That is too blunt, and it makes people walk away from deals worth taking.
- Amazon dominates
- Holds the buy box almost permanently, never runs out of stock, undercuts third parties. Walk away.
- Amazon price matches
- Shares the buy box with sellers who match its price. Workable, provided your maths hold at Amazon's price rather than the list price.
- Amazon runs out regularly
- Gaps of two or three weeks appear in the history. This is an opportunity: experienced sellers hold stock specifically to sell into those gaps at a premium.
Telling them apart needs the buy box history, which is why a scanning tool that just shows a red flag is throwing away the useful part of the answer.
How to source efficiently
A productive sourcing trip is a numbers game. You might scan two hundred items and buy six. The skill is rejecting quickly, so the rate of scanning matters as much as the quality of the decision.
What separates a good trip from a wasted one
- Work the clearance ends first. That is where the gap between shelf price and Amazon price is widest.
- Scan everything, not just what looks promising. Intuition about what sells well on Amazon is usually wrong.
- Reject fast. If the eligibility check fails, move on without reading the profit.
- Note where you were. Which shop, which aisle, how many were left. You will want to go back.
- Check how many are on the shelf. Six units of a slow seller is months of tied-up capital.
- Keep a record. Scanning the same product on three separate trips because you cannot remember rejecting it is wasted time.
AskJeffy saves every scan with the shop and your notes, so a week later you can see what you looked at, what you bought, and what you paid. Scan the same product months on and it tells you what you concluded last time.
What you need to start
Less than most people assume. An Amazon seller account, a phone, and enough capital to buy stock you can afford to be wrong about.
- A seller account. The Individual plan costs nothing monthly but charges per sale, which is fine for testing.
- A scanning app. You cannot make these decisions by eye, and checking each product manually on Amazon is far too slow.
- Starting capital. A few hundred pounds is enough to learn with. Our startup costs guide covers the numbers.
- Somewhere to prep. Boxes, labels and a printer, or a prep service if you would rather not handle stock.
The real constraint is not money. It is the discipline to reject nine out of ten products that look fine but are not, and to keep walking when the numbers do not work.
Scanning a barcode with AskJeffy
AskJeffy answers all four questions from a single scan, on iPhone and iPad.
Point the camera at any barcode and the app matches it to the Amazon listing. The checks that stop a purchase come first, because that is how the decision actually runs. Then the sales, counted from rank drops and shown as a position within the category. Then the money, led by the most you can pay. Then the risk: price history, seller counts, and a warning when a collapse looks likely.
There is no score out of a hundred. You see the checks and what each one found, so every line can be verified against the listing in front of you. Where the data is too thin to judge, it says so rather than guessing.
What else it does
- Saves every scan, with the shop, your notes and what you paid.
- Records what you bought and how many, so the list doubles as a buy list.
- Shows what a hundred units means: total outlay, total return, and how long it takes to shift.
- Exports a day of sourcing to a spreadsheet.
- Tells you what you paid last time when you scan something again.
The scanner is live on iPhone and iPad. Start free and scan your first product today. Try AskJeffy free
Frequently asked questions
What is retail arbitrage on Amazon?
Retail arbitrage means buying products cheaply from a shop, usually on clearance or promotion, and reselling them on Amazon at a higher price. You list against the existing product page rather than creating your own listing, so there is no supplier, brand or minimum order to arrange.
Is retail arbitrage still profitable?
Yes, but margins are thinner and competition is faster than a few years ago. The sellers who do well reject most of what they scan, check price history before buying, and avoid listings where sellers are arriving faster than the product sells.
How much money do you need to start retail arbitrage?
A few hundred pounds is enough to learn with, since you buy in small quantities and sell within weeks. The main costs are stock, Amazon fees and a scanning app. There is no minimum order to meet and no product development to fund.
What ROI should I aim for in retail arbitrage?
Around 30 percent is the usual baseline, but it varies. Items under five pounds usually need 50 to 100 percent because the handling cost is the same as an expensive item. High ticket items above seventy five pounds can work at 18 to 25 percent. Q4 toys need 45 percent or more to absorb the post-holiday price drop.
Do I need an app to do retail arbitrage?
In practice yes. Checking each product manually on Amazon takes minutes, and a sourcing trip may involve scanning two hundred items. A scanning app resolves a barcode to the listing and shows the fees, sales and competition in seconds.
How do I know if I am allowed to sell a product?
Amazon restricts some categories and some brands. Tools can flag categories that usually need approval and brands known to restrict resellers, but no third-party app can see your own account approvals without connecting to your seller account. Confirm in Seller Central before buying stock.
Why do prices collapse on arbitrage listings?
When a good clearance deal is discovered widely, many sellers buy the same stock and list it at once. Supply overwhelms what the listing normally sells, and sellers undercut each other to move it. Checking whether seller numbers are rising while the price falls will usually warn you before you buy in.
What is the difference between retail arbitrage and online arbitrage?
Retail arbitrage means buying from physical shops, and online arbitrage means buying from other retailers' websites. The buying decision is identical: the same checks on eligibility, sales speed, fees and price stability apply to both.