Skip to main content

Chapter 8: BSR, sales and demand

What you'll learn: what Best Sellers Rank tells you, why a "good" rank depends on the category, how ArbiSource's BSR % makes ranks comparable, and how to tell an estimate of sales from a confirmed figure.

What BSR is​

Every product that has sold on Amazon has a Best Sellers Rank (BSR) in its category. A rank of 1 means it's currently the best-selling product in that category. The lower the number, the more it's selling.

A few things to know about it:

  • It's relative. BSR compares a product's recent sales with every other product in the same category. It doesn't tell you a number of sales on its own.
  • It moves all the time. When a unit sells, the rank usually improves (the number drops). Between sales it slowly drifts upwards. A single snapshot can be misleading, so always look at the history (Chapter 9).
  • It's per category. A product can have a main category rank and one or more sub-category ranks. ArbiSource uses the main category rank.

Why a "good" rank depends on the category​

Categories are very different sizes. A rank of 50,000 could be a steady seller in a huge category such as Home & Kitchen, and a slow one in a much smaller category. So a single BSR rule of thumb ("anything under X") doesn't work across categories.

BSR %: comparing like with like​

ArbiSource solves this with BSR %: the product's rank divided by the number of ranked products in its category, shown as a percentage. A rank of 5,000 in a category of 500,000 ranked products is 1%, meaning the product is in the top 1% of its category, however big that category is. The lower the percentage, the better it sells relative to its category.

That makes BSR % the better choice for filters. It works the same way in every category, so one setting does the job everywhere. In your filters, use BSR % rather than a raw BSR, and leave it fully open at first (4.3).

Minimum or maximum?

A maximum BSR (or BSR %) hides slower sellers, which is usually what you want. A minimum BSR hides the fastest sellers, because they have the lowest numbers. If you ever set a minimum by mistake, you can lose your best results. It's one of the most common reasons for "no results".

A UK result panel header with the BSR and its BSR % in brackets highlighted (for example BSR: 19,198 (0.31%)), plus the Sales Velocity section of the Filters drawer with the BSR % | BSR # toggle and the BSR % slider fully open (0 to 30%+) (1 of 2) A UK result panel header with the BSR and its BSR % in brackets highlighted (for example BSR: 19,198 (0.31%)), plus the Sales Velocity section of the Filters drawer with the BSR % | BSR # toggle and the BSR % slider fully open (0 to 30%+) (2 of 2)

Estimated sales and confirmed sales​

Amazon doesn't publish how many units a product sells. So there are two kinds of sales figure:

Where it comes fromHow to use it
Estimated monthly salesWorked out from the product's rank and how often its BSR dropsA good guide to demand, but an estimate. Treat it as a range, not an exact number
Confirmed salesAmazon's own "bought in past month" figure, where Amazon shows one on the listingStronger evidence of demand. Not every listing has one

ArbiSource shows both on each result, with confirmed sales from the past 30 days where available. When the two broadly agree, you can be more confident.

Your share of the sales​

A product selling 100 a month doesn't mean you will sell 100 a month. Sales are shared between the sellers who win the Buy Box (Chapter 10). As a rough way to think about it, divide the monthly sales by the number of competitive FBA sellers, then be a little cautious. That gives you a sensible idea of how many units you could sell in a month, and so how many to buy.

Example

a product sells about 60 units a month, and four FBA sellers are priced close to the Buy Box. Your share might be around 15 a month, so a first order of 5 to 10 units would sell through in a few weeks, leaving room to buy more if it goes well.

Steady demand beats a spike​

What you're looking for is regular sales over months, not one busy week. When you check a product, ask:

  • Has it been selling steadily over the last 90 to 180 days?
  • Is demand seasonal? Toys, gifts and garden products can sell far more at certain times of year. Look back 12 months (Chapter 9).
  • Is the rank improving, steady or getting worse over time?

ArbiSource's result panel helps here too. Its past-performance table shows the BSR (rank and BSR %) and the number of BSR Drops for the last 30, 90, 180 and 365 days, alongside the average Sell Price and Profit / ROI, so you can see at a glance whether a product sells consistently.

Estimated and confirmed sales (from 4:39), then average rank, BSR % and BSR drops (from about 5:20). From "Mastering Result Panels" (ArbiSource), recorded on an earlier version of the interface and in dollars; the figures work the same way.

The past-performance table in a UK result panel (columns Sell Price, Profit / ROI, BSR as rank / %, BSR Drops; rows 30-Day, 90-Day, 180-Day, 365-Day)

Checklist​

  • I know a lower BSR means more sales, and that it only means something within its category
  • I use BSR % for filters, and I never set a minimum BSR by accident
  • I know the difference between estimated and confirmed sales
  • I think about my share of the sales before deciding how many to buy