Analyzing Products With the Keepa App

Keepa records the price and sales rank history of Amazon products. The mobile app puts that history in your hand at the shelf, which is what turns a scan into a judgement. This lesson covers setting it up and reading it quickly.

Set it up first

Before scanning anything, open Settings in the app and configure it once:

• Choose your marketplace. The UK site, the US site, or wherever you sell. Reading a US chart for a UK purchase is a straightforward way to buy the wrong thing.
• Choose which lines to display. The four that matter for arbitrage are sales rank, Buy Box, New and third-party FBA. Turn the rest off; a chart with ten overlapping lines is unreadable on a phone.
• Set the default range. 90 days is a sensible working default: long enough to show a pattern, short enough that it reflects the current state of the listing.

Scanning

Go to Search and either type a product name or tap Scan and read the barcode. The chart loads for that product.

What to look at, in order

Is Amazon selling it

Check first, because it can end the decision. If Amazon itself is on the listing, competing is difficult: Amazon usually holds the Buy Box, and your stock may sit unsold. Where there is no Buy Box line at all, confirm that Amazon is not the seller before going further.

Is the sales rank consistent

This is the question the sales rank lesson said you could not answer from a single reading, and here is where you answer it. You want the average rank to be stable across periods rather than jumping about.

The example in the video shows an average of 58,000 over 60 days and about 40,000 over 180 days. Two similar numbers over different windows mean the product sells steadily. A recent average far better than the longer one usually means something temporary happened, and it will revert.

How many drops

This is the most useful single number Keepa gives you. Each significant drop in sales rank indicates a sale, so the drop count estimates how often the product sells.

The example shows 32 drops in 30 days, 101 in 90 days and 223 in 180 days. That is roughly 32 sales a month, and the three figures are consistent with each other, which means the rate is real rather than a spike.

Now you can do the arithmetic that actually decides the purchase. If a product sells around 32 times a month and there are 8 sellers sharing it, you can expect roughly 4 sales a month. Buy 20 units and you are holding stock for five months. Buy 4 and you sell out in one. The drop count plus the seller count is what tells you how many units to buy, and that is the question most new sellers never ask.

Is the price consistent

Look at the Buy Box and New lines. A steady price means the margin you calculated is likely to hold. A price sliding downwards over months means your margin will be smaller by the time you are selling. A price with sharp spikes means the good number you are looking at may be one of the spikes.

What Keepa does not tell you

The seller count. Keepa shows the price and rank history, not how many people you are sharing the listing with, and without that the drop count cannot be turned into your own expected sales.

Get it from the Amazon Seller app or from SellerAmp SAS, both covered in the previous two lessons. In practice most sellers either use SellerAmp, which shows the chart and the competition together, or run Keepa alongside the free Seller app.

The checklist at the shelf

1. Is Amazon on the listing? If yes, usually walk away.
2. Is the average rank consistent across 90 and 180 days?
3. How many drops, and therefore how many sales a month?
4. Divide by the number of sellers. Is that rate acceptable to you?
5. Is the price stable enough that the margin will still exist when you sell?

That takes under a minute once it is familiar. The Advanced Keepa Sourcing module later in the course goes through the chart properly, line by line.