asinIQAmazon tells you how many people bought a product last month. It doesn't tell you how many will buy from you once you join the listing. That second number decides whether a wholesale order turns a profit or sits in a warehouse, so it's worth five minutes of math before you buy.
The short answer
To estimate your sales on a shared Amazon listing, take the "bought in past month" number for the exact variation you'd sell. Divide it by the number of sellers priced close to the Buy Box, counting yourself. Then plan on about two-thirds of that.
Here, four sellers are priced near the Buy Box, so you'd be the fifth. If Amazon itself sells on the listing, start from about a quarter of the badge number instead.
Try it with your product
Read these off the product page. Nothing you type leaves this page.
An estimate, not a promise. The badge counts customers, so someone who buys several units counts once.
The two-thirds comes from 90 days of Buy Box records on 1,464 Amazon.com and Amazon.ca listings where four or more sellers took turns. On those listings, the seller in the middle of the pack got a median of 66% of what an even split would have given them. On listings where Amazon and other sellers both won the Buy Box, Amazon held it a median 67% of the time.
Start from the right sales number
The line under the price that reads "100+ bought in past month" is the most direct sales figure Amazon publishes. An Amazon staff member explained on the Seller Forums how it's counted:
- It counts unique customers over the past 30 days. A customer who buys three units counts once.
- It needs at least 50 customers before it appears.
- It runs 3 to 4 days behind.
- The product must have been for sale for at least 30 days, and some categories never show it.
Amazon also rounds the "bought in past month" number down to steps: 50+, 100+, 200+ and so on. "100+" can mean anything from 100 to 199 customers. Use the number exactly as shown. Because it's a floor, your estimate stays cautious.
If the variation you want shows no badge, fewer than 50 customers bought it in the past month, or the listing is too new, or its category doesn't show the badge. Take extra care with listings that lack the badge, because without it you're estimating sales from indirect signals.
Make sure the number belongs to your variation
Many listings are a family of variations: sizes, colors, pack counts. Each variation has its own product code (ASIN), and you would sell one of them. Nearly two in three of the products we looked at were one variation of a larger family, so this step comes up often.
It's tempting to treat the badge as a total for the whole family and divide it again by the number of variations. In our data, the badge belongs to one variation.
The figure you see is for the variation you have selected. Click each variation and read its own number before you divide anything.
Reviews work differently. Amazon usually pools reviews across colors, sizes and pack counts, so the review count on a page often covers the whole family. Since February 2026, Amazon has stopped sharing reviews between variations that work differently, such as two flavors of a protein powder.
When your variation has no badge and you still want a rough split, open the reviews and look at the variation named on each recent one. The share that mention your variation is a rough guide to its share of sales. Treat it as a last resort. Reviews trail sales by weeks, and buyers of some variations leave more reviews than others.
Count the sellers you'll actually compete with
The Buy Box is the "Add to Cart" button. Only one seller holds it at a time, and nearly every sale goes through it. Amazon doesn't publish exactly how it chooses the winner. Price, delivery speed and the seller's track record all count, and sellers priced close together take turns.
A seller listed at twice the Buy Box price rarely gets a turn, and neither does one with slow delivery. Open the list of offers and count only the sellers priced close to the Buy Box. Then add one for yourself. Your arrival splits the sales one more way, and an estimate that leaves you out is too high from the start.
Expect less than an even split
Sellers who share a listing don't share it equally. The busiest seller usually takes much more than an equal slice, and the rest split what's left.
| Sellers taking turns | Listings | Even split | Top seller | Middle seller |
|---|---|---|---|---|
| 2 or 3 | 1,473 | 33% to 50% | 73% | 41% |
| 4 or 5 | 872 | 20% to 25% | 50% | 17% |
| 6 to 9 | 522 | 11% to 17% | 38% | 8.5% |
| 10 or more | 70 | 10% or less | 29% | 5% |
On Amazon listings where four or five sellers took turns at the Buy Box, the top seller held the Buy Box more than half the time on 49% of them. The middle seller got a median 17%, about two-thirds of an even split. The pattern was much the same on Amazon.com and Amazon.ca.
These figures count only sellers who held the Buy Box at least 1% of the time. Sellers who listed and never won it aren't in the table, so a brand-new seller can do worse than the middle seller.
When Amazon sells on the listing
When Amazon is one of the sellers, it takes most of the sales. On 1,313 listings where Amazon and at least one other seller won the Buy Box over 90 days, Amazon held it a median 67% of the time. All the other sellers together held about 28%.
When Amazon sells on the listing, start from about a quarter of the "bought in past month" number, then divide it by the sellers near the Buy Box, counting yourself, and take two-thirds. With a 100+ badge, the other sellers share about 28 of those sales. Divided between five sellers that's 5.6, and two-thirds of 5.6 is about 4 sales a month. Four a month rarely justifies a wholesale order.
How many units to buy first
Turn your estimate into a first order. About a month of your estimated sales is a sensible test. It's enough to see real sales, and small enough that a wrong guess is cheap. Keep testing new products a few units at a time. Once the first units sell, reorder from what actually sold instead of from the estimate.
The same math tells you when a supplier's minimum order doesn't fit.
That's more than three years of stock, paying storage fees every month, while new sellers can push the price down. Ask for a smaller opening order.
What can raise or lower your share
Start with price. An offer priced well above the Buy Box rarely wins it, and cutting below everyone else can start a race down. Watch for new sellers on the listing. A seller who doesn't yet know their fees may keep lowering the price just to get sales.
Staying in stock matters too. A seller who runs out gets no turns until they restock, and the others split the sales in the meantime.
Sales also change with the season, and the "bought in past month" badge only covers the last 30 days. A product that sold 100 in December can sell 40 in February. Look at how the price and the number of sellers have moved over several months before you commit to a large order.
Common questions
Does "bought in past month" count units or customers?
Customers. An Amazon staff member said on the Seller Forums that the badge counts unique customers who bought in the past 30 days. It needs at least 50 of them to appear, and it runs 3 to 4 days behind. A customer who buys several units counts once, so units sold can be higher than the badge shows.
Is the "bought in past month" number for the whole listing or one variation?
For the variation you're looking at. Of 194 variation families we tracked where more than one variation showed the badge, 161 (83%) showed different numbers on different variations.
How many sellers is too many on an Amazon listing?
There's no fixed number. Ten sellers on a variation that sells 1,000 a month still leaves each of them a useful share. Five sellers on one that sells 50 a month doesn't. Work out your own share: the badge number for your variation, divided by the sellers near the Buy Box price plus you, times two-thirds. Then check whether that many sales pays for the order.
How much of the Buy Box does Amazon take when it sells on a listing?
In our data, a median 67% of the time on listings where Amazon and other sellers both won the Buy Box over 90 days. All the other sellers together held about 28%.
If I match the Buy Box price, will I get an equal share of sales?
Usually less. On listings where four or more sellers took turns, the seller in the middle got a median two-thirds of an even split. The top seller often held the Buy Box more than half the time.
About the data
We checked 10,169 products between 9 and 21 October 2026: 2,843 on Amazon.com and 7,326 on Amazon.ca. Each record included 90 days of Buy Box history, showing which sellers held the Buy Box and for what share of the time.
We kept the 7,698 listings where some seller held the Buy Box for at least half of those 90 days. Of those, 2,497 had Amazon in the rotation. The 1,313 where Amazon and other sellers both won are the basis of the Amazon section. The remaining 5,201 listings, without Amazon, make up the table. A seller counts as taking turns if they held the Buy Box at least 1% of the time. The products lean toward what resellers research (median sales rank about 3,000), so read these as figures for products in a typical sourcing range, not for all of Amazon.