How Amazon price drops work, and how to catch them

Every online arbitrage buy starts the same way: a price on Amazon moved, and it moved far enough that buying the item and selling it back on the same marketplace, or a different one, leaves you with money. That is the whole trade. The hard part is not understanding it. The hard part is that Amazon UK prices change constantly, most of the changes mean nothing, and the handful that do mean something are gone within hours.

This is a walk through what actually makes an Amazon price move, why the size of a drop tells you far less than people assume, and the realistic ways a side-hustle seller catches the ones worth having without spending every evening refreshing a deals page.

The four things that move an Amazon price

"Price drop" gets used as though it were one event. It is at least four different ones, and they behave differently enough that treating them the same is the first mistake.

• Amazon repricing its own offer. When Amazon is a seller on the listing, its algorithm moves the price on its own schedule: overstock, a competitor matching elsewhere, a supplier deal. These are the deepest drops and the least predictable, and because Amazon holds the Buy Box more often than not, an Amazon-side drop usually drags the visible price down with it.
• Third-party sellers repricing against each other. Most listings have several sellers running automated repricers pointed at each other. A crowded listing can walk itself down several percent in an afternoon with nobody deciding anything. These drops are real, but they are also the ones most likely to reverse, because the same repricers walk the price back up when stock thins.
• Limited-time deals. A scheduled promotion with a start and an end. The discount is genuine and the deadline is genuine, which makes these the most actionable of the four, and also the most competitive, because every other seller with an alert set can see the same thing.
• One-time coupons on the listing. Not a price change at all. The listing price stays where it is and a tickable coupon comes off at checkout. This matters enormously for arbitrage, because it means your actual buy price is lower than the price any tool showing the headline figure will tell you.

There is a fifth kind that belongs to a different trade entirely: a retailer dropping its own price, in a supermarket or on a retailer's website, with no relationship to what the item sells for on Amazon. That gap is where retail arbitrage lives, and it behaves nothing like an Amazon-side drop. Worth keeping the two separate in your head, because tools, alerts and habits that work for one are close to useless for the other.

Why the percentage is the wrong number to look at

The instinct is to sort by biggest discount. It is almost always wrong, for three reasons.

A percentage is measured against something, and that something is usually not the real price. A drop from an inflated reference price is not a drop. A listing that sat at £24.99 for a fortnight and briefly touched £39.99 last month can advertise a 37 percent saving that never existed for a buyer. What you want is the price relative to where the item has actually traded, which is a question about history, not about today.

A drop on a product nobody buys is not an opportunity. You can find 60 percent off all day on lines that sell twice a year. The discount is real and the money is dead, because your capital sits in a box for eleven months and then pays storage on itself. Sales velocity is not a nice-to-have next to the discount, it is the thing that decides whether the discount is worth anything.

And the drop tells you nothing about your side of the sum. A 30 percent drop on a low-margin category with a heavy fulfilment fee can still be a loss. The referral fee, the fulfilment fee, storage, and the digital services fee all come out before you see a penny, and several of those step at thresholds rather than scaling smoothly. Our walk through every fee behind an Amazon FBA calculator covers where that arithmetic goes wrong most often.

The three questions a drop has to survive

Before a drop is a buy, it has to answer three things, in this order, because each one can kill it on its own.

1. Does it clear after fees? Buy price, including any coupon, against the price it realistically sells at, minus every Amazon fee and your own prep and shipping. Not the headline saving. The number at the end.
2. Will it actually sell? Monthly sales and how many sellers are already on the listing. Profit divided by nothing is nothing, and a listing with twenty sellers means your share of those sales is small even if it moves well.
3. Are you allowed to list it? Plenty of brands and categories are restricted on Amazon UK, and finding out after the stock arrives is an expensive way to learn. Connect your Amazon selling account to retailscout and it checks whether you can list a product before you buy, which is the check people skip and then regret.

How people try to catch drops by hand, and where it breaks

The manual methods all work. They just do not scale past a handful of products, and the failure is always the same shape: you are trading your evenings for coverage.

• Watching the deals pages. Fine for browsing, hopeless as a system. The pages are ranked for shoppers, not resellers, so what you see is what Amazon wants to sell rather than what has moved against its own history.
• Per-product price alerts. Genuinely useful, and the right tool if you already know the twenty products you want. It cannot find you something you have never heard of, which is most of what is worth buying.
• Spreadsheets of ASINs you check on a Sunday. The list rots. Prices moved on Tuesday and reverted on Thursday, and by Sunday the only thing your list proves is that you were not looking when it mattered.

The common thread is that catching drops is a filtering problem, not a looking problem. There is no shortage of price movement on Amazon UK. There is a shortage of attention, and every hour you spend scrolling is an hour you did not spend buying.

What a drop recipe actually looks like

This is the part worth stealing whether or not you ever pay for a tool: a useful drop filter is a set of conditions, all of which have to be true at once, and none of which is the discount on its own.

For a concrete example, here is what retailscout's own sweeps ask for. There are nine recipes, eight of them running across Amazon UK and one comparing the European marketplaces against the UK sell price. These are the recipe definitions as they stand on 9 September 2026:

• Fresh price drops - the Buy Box and Amazon's own price both down 20 percent or more in the last day, at least 2 sellers on the listing, £10 or more, newest change first.
• Fast movers - limited-time deals down 15 percent or more in a day, 500 or more sold a month, priced £10 to £25, with 4 to 10 sellers.
• Grocery drops - down 20 percent or more in the last day, 200 or more sold a month, 3 or more sellers, £10 or more.
• EU to UK - Amazon Germany, France, Italy or Spain down 10 percent or more in the last day against the UK sell price, 50 or more sold a month, 2 or more sellers.

Look at what those have in common. Every one pairs a movement condition with a demand condition and a competition condition, and every one has a price floor, because pennies of margin on a £4 item is not a business. The seller counts are doing real work too: the fresh drops recipe asks for only 2 sellers precisely because it also requires Amazon to have dropped, so 2 sellers means Amazon plus one other, which is a thin, low-competition listing rather than a dying one.

You can build filters in that shape yourself with any decent product database. The reason to have something run them for you is simply that a condition set is worthless if it only runs when you remember to run it.

The trade-off nobody tells beginners about

Velocity and margin tend to select against each other, and it is structural rather than bad luck. The high-volume lines on Amazon UK are the ones every seller can see, so they are competitively priced, so the margin on them is thin. The fat margins sit on slower lines, or on clearance in a shop where the retailer wants the shelf back and is not thinking about what the item fetches on Amazon.

That is not an argument against catching drops. Thinner margin at higher volume with no petrol and no Saturday spent walking aisles is a perfectly good way to trade, and it is how a lot of sellers do most of their turnover. But it does mean you should be suspicious of anyone showing you triple-digit returns from a drop feed, and it means the honest answer to "which is better, online or in store" is that they are different trades with different economics and most established sellers run both.

Set your minimum ROI accordingly, and set it once. A filter that only shows you what clears your own threshold is the single biggest saver of time in online arbitrage, because it removes the temptation to talk yourself into a marginal deal at eleven at night.

The three things people call a "voucher"

Worth untangling, because merging them is how buy prices end up wrong.

• An Amazon one-time coupon, ticked on the listing and taken off at checkout. It changes what you pay, so it belongs in the buy price. retailscout has two sweep recipes dedicated to these, and both subtract the coupon before working out profit, which is the only way the ROI you see is the ROI you get.
• A retailer's own checkout code, from a newsletter or a promotion on the retailer's site. Nothing to do with Amazon. These are harvested separately in retailscout, with expiry countdowns, and they show up on the retailer's own site through the browser extension.
• An on-listing promotion, such as multibuy or a percentage off at a threshold. Real money, awkward to model, and the thing most likely to make a manual calculation disagree with your actual invoice.

Where retailscout fits

retailscout's online arbitrage suite runs those recipes for you and refreshes them through the day, so the drops arrive already priced against the Amazon UK sell side, with fees, ROI and monthly sales attached, and filtered to your own minimum ROI rather than to a generic one. Alongside the feeds there are radars you point at the brands and shops you actually sell, daily sweeps on watched brands and storefronts, twice-a-day checks on retailer sale pages, and a watchlist that keeps an eye on the deals you could not buy today.

It is one membership at £14.99 a month, with a 7-day free trial and cancel anytime, and it includes the deal analyser and the in-store trip planning rather than charging separately for each.

The part worth saying plainly: this is sourcing you drive, not a list of picks posted to everyone at once. What the feeds surface depends on the brands, shops and thresholds you point them at, which is a different product from a shared leads list, and it is the difference that decides whether you are competing with every other subscriber for the same twenty deals.

The takeaway

Amazon prices move all day, and almost none of it matters. What matters is a movement that is large against the item's own history, on something that sells, on a listing you are not fighting twenty other sellers for, in a category whose fees leave you something, on a product you are allowed to list. That is five conditions, and any one of them failing is a bad buy dressed as a bargain.

Write those conditions down as a filter, run them on a schedule rather than on a whim, and check the fee arithmetic before the money leaves rather than after. If you are still working out the basics of the model, start with our guide to starting online arbitrage in the UK, which covers the beginner traps in more detail.