How does retail arbitrage work?

Short answer: you buy new products at retail discounts — clearance racks, markdowns, end-of-season sales — and resell them on Amazon for more. The gap between the two prices is your business, and Amazon's fees decide how much of it you keep.

Retail arbitrage is one of the oldest ideas in commerce wearing modern clothes: buy low in one place, sell high in another. What makes it work in 2026 is a smartphone. You walk the clearance aisle, point your phone at a barcode, and instantly see what the item sells for on Amazon, what the fees would be, and whether the trip was worth it. What used to require instinct and research now takes three seconds.

It is also harder than it looks, thinner-margined than the gurus admit, and shaped by rule changes that keep moving the goalposts. Here is the full picture.

The basic loop

The mechanics are straightforward enough to explain in a paragraph.

You visit retail stores — big-box chains, discount outlets, pharmacies, toy stores — and head for the clearance sections. You scan product barcodes with the free Amazon Seller app, which shows the current Amazon price, the sales rank (how fast it sells), whether your account is allowed to sell that brand, and an estimate of fees and profit. When the numbers work, you buy. You prep and label the products — since January 2026, Amazon requires all FBA inventory to arrive fully prepped, so this step is on you — ship them to an Amazon warehouse, and Amazon handles storage, customer service, and delivery when orders come in.

That is the whole loop: source, scan, buy, prep, ship, get paid. About a quarter of Amazon sellers use some form of retail arbitrage, and the average practitioner generates roughly $11,600 a month in sales at around 21% margins — about $2,400 a month in profit. A third of them started with less than $500 in capital. The barrier to entry is among the lowest in e-commerce: a phone, a scanning app, and willingness to walk stores.

The simplicity is real. So is everything below.

The math, worked honestly

This is where most beginners get hurt, because they buy based on the gap between the store price and the Amazon price without accounting for everything in between. Here is a realistic breakdown.

Say you find a product on clearance for $18. It sells on Amazon for $39.99. The spread looks like $22. Now subtract:

• Product cost: $18
• Referral fee: about $6 (Amazon takes 8–15% depending on category, up to 45% in some)
• FBA fulfillment fee: about $5.25 (picking, packing, shipping to the customer)
• Prep and inbound shipping: about $2 (labels, poly bags, getting it to the warehouse)
• Return allocation: about $1 (a share of expected returns, higher in some categories)

Net profit: $39.99 − ($18 + $6 + $5.25 + $2 + $1) = $7.74 per unit. The $22 spread became $7.74 — a margin under 20%.

That is a normal, decent find. Notice how quickly the fees consumed the gap: Amazon and logistics took nearly two-thirds of the spread. This is why experienced sellers target a minimum of 50% return on investment — buy at $10, net at least $5 after everything — and why anything below 30% ROI is generally not worth the shelf space. The fees are not the enemy; underestimating them is.

The tools that make it possible

Three tools separate systematic sellers from hopeful ones.

The Amazon Seller app (free) is the foundation. Scanning a barcode shows the live Amazon price, your eligibility to sell the item, estimated fees, and approximate profit at the scanned price. It is the three-second decision engine the whole model runs on.

Keepa tracks Amazon price and sales-rank history. This matters because a product selling for $30 today might have been $15 last week — a temporary spike that makes a losing buy look profitable. Keepa shows you the history so you buy based on the real price, not the momentary one. Sellers who skip this step learn the lesson through dead inventory.

BrickSeek verifies that the clearance price you see is actually available at nearby stores before you drive across town. Nothing wastes an afternoon like hunting a deal that existed only at one location, yesterday.

None of these tools find deals for you. They verify deals you find. The finding — the store visits, the pattern recognition, the knowledge of which categories clear out when — is the actual skill, and it only comes from repetition.

What changed in 2026

Retail arbitrage in 2026 is not the same game it was in 2023, and the differences all squeeze margins.

Tariffs reshaped sourcing. Multi-layered tariffs on imports, some reaching very high rates, largely killed international online arbitrage — buying from overseas suppliers and flipping on Amazon no longer pencils out for most categories. Domestic retail sourcing, the clearance-aisle kind, was less affected, which perversely made it relatively more attractive.

Amazon ended FBA prep services. Since January 2026, products must arrive at warehouses fully prepped and labeled. Sellers either do it themselves — five to fifteen minutes per unit — or pay a prep center $0.30 to $1.00 per unit. Either way, it is a new cost line that did not exist before.

Fees kept compressing. Referral fees now span 8–45% by category, fulfillment fees rose, and storage surcharges penalize slow-moving inventory. The categories where sellers once scraped by on 15% margins no longer work; you need 25% or more before fees to keep anything meaningful.

Competition intensified. More sellers scanning the same clearance aisles means faster price erosion once multiple sellers list the same item. A profitable listing with ten sellers becomes a breakeven listing within weeks. Speed of sourcing and listing is a genuine competitive advantage.

None of this killed the model. But the era of scanning everything in sight and profiting is over. What works now is operational discipline: tighter ROI thresholds, faster turnover, better category selection.

The unglamorous parts

A few realities that do not appear in the highlight reels.

Gated brands and categories. Some brands require approval before you can list their products. Buy $200 of inventory you cannot list and you own a very expensive lesson. Always check eligibility in the seller app before buying in bulk — the app flags this, but only if you look.

No control over supply. You are buying from retailers, not wholesalers. When the sale ends, it ends. When the stock is gone, it is gone. Find a winner and you can only ride it until the clearance rack empties, and by the time the deal returns — if it ever does — your customers may have moved on. This is the structural ceiling of the model: you cannot scale a product, only a process.

Returns and storage. Return rates vary by category and eat directly into margins. Slow-moving inventory accrues monthly storage fees and, eventually, long-term surcharges. Turnover is not just nice to have; it is the difference between a business and a warehouse bill.

It is a job. Sourcing means driving to stores, regularly, including the unglamorous ones. Prep means labeling boxes at your kitchen table. The sellers making real money treat it as operations — routes, schedules, systems — not as treasure hunting. The treasure-hunting framing is marketing. The operations framing is the business.

The first-week plan

If the model appeals to you, here is a sane way to test it without committing your savings.

Day one: set up. Create an Amazon seller account — the individual plan has no monthly fee, you pay per item sold — install the Seller app, and add Keepa's browser extension. Read the restricted-products list for your country so you know what you cannot touch before you fall in love with a deal.

Days two to four: scan, don't buy. Visit three or four stores and scan everything in the clearance sections. Buy nothing. You are calibrating — learning which categories show real spreads, what sales ranks actually move, and how often the app's profit estimate survives Keepa's history check. Most beginners who skip this step buy their education in the form of dead inventory.

Days five to seven: one small shipment. Pick your five best finds — high ROI, fast sales rank, ungated brands — buy modest quantities, prep and label them properly, and send a single box to FBA. Then watch: how fast they sell, what the real fees look like, how returns behave.

One week, under $200, and you will know more about this business than any guide can teach — including whether you actually enjoy it. That last part matters more than people admit. The sellers who last are not the ones who found the best first deal. They are the ones who did not mind the driving.

Who it suits

Retail arbitrage suits a specific profile: someone with more time than capital, comfortable with physical legwork, disciplined about numbers, and realistic about scale. It is an excellent first e-commerce education — you learn pricing, fees, sales rank, and cash flow with $500 instead of $50,000. Many successful Amazon sellers started exactly here and graduated to wholesale or private label once they understood the platform.

It does not suit anyone looking for passive income, anyone who hates driving to stores, or anyone unwilling to do arithmetic before every purchase. The math is the job. If running the numbers feels like a chore, the model will feel like a chore, because they are the same thing.

There is a quiet honesty to retail arbitrage that the more glamorous online businesses lack. No one is promising you freedom or leverage or 10x. The proposition is plain: find things priced wrong, capture the difference, keep your costs tight. In a corner of the internet built on exaggerated promises, a business that tells the truth about its own margins is almost refreshing. Buy low, sell high, count everything — it worked a century ago, and with enough discipline, it still works now.