Lesson 5 of 7

Amazon Product Ideas Research

Working backwards from Amazon listings to the brands worth approaching, using reverse search.

This is reverse search applied to wholesale: find products worth selling on Amazon first, then go looking for whoever supplies them. It is the most systematic route to a supplier, because it starts from demand you have verified rather than from a catalogue somebody sent you.

The same process finds online arbitrage deals. The only difference is what you do with the list at the end: for wholesale you search the product name plus "wholesale", for arbitrage you search the product name and look for a cheaper retailer.

Six ways to generate candidates

Your own inventory. The favourite, for the reason given in the previous lesson: you already know these products sell.
Related products. Branching out from a product you like into what Amazon shows alongside it.
Brand search. Finding one good product from a brand and going through the rest of the range.
Keyword search. Searching for what you want to sell.
Category browsing. Working down through Amazon's category tree.
Competitors. Looking at what other sellers stock. The second favourite, and covered in detail below.

The wholesale criteria, and how they differ from arbitrage

Most of the checks are the same as arbitrage. Two are different, and the difference matters.

No Amazon on the listing. Same as always.
More than three FBA sellers. This one is reversed from what you might expect. In arbitrage, fewer competitors is better. Here, too few sellers means the product is probably private label. One or two FBA offers usually means a brand owner selling their own product, sometimes with a second listing under their own name, and there is no supply route into it for you. Three or more sellers is evidence that the product is genuinely distributed.
Price at least 15 in your currency. See the arithmetic below.
Consistent price and sales rank, checked over 90 days as in the Keepa module.
No more than two or three competitors near the Buy Box. Note this is different from the total seller count. Ten FBA sellers on a listing is fine if only two or three are priced near the Buy Box; the rest are not competing with you for the sale.
At least 15 to 20 sales a month.

Why 15 is the price floor

The reasoning is worth following, because it gives you a rule you can apply at a glance.

A common rough shape for a wholesale deal is thirds: about a third of the selling price to buy the product, about a third to Amazon in fees, about a third left. On a product selling at 15, that is roughly 5 to buy, 5 in fees and 5 kept.

Below 15, the arithmetic stops working, because Amazon's fees do not shrink proportionally on cheap items. So 15 is the point at which a normal wholesale margin still leaves something worth handling.

Used in reverse, this is a fast filter: divide the Amazon price by three and that is roughly what you need to buy at. If no wholesaler will sell it to you near that figure, move on.

Working a category

Start with a category, drill into a subcategory, and use the filters. Set a price range matching your criteria, such as 15 to 50. Hover for the Keepa chart to dismiss anything Amazon is selling.

Doing this by hand is slow, so a research tool that filters listings in bulk speeds it up considerably. The lesson uses Jungle Scout, with filters for: no Amazon on the listing, Buy Box held by FBA or FBM sellers, a minimum price, a sales rank ceiling, and at least three sellers.

The rank ceiling used is 150,000 in Toys, with the reasoning stated plainly: at that rank the stock sells more slowly, in the region of a couple of months rather than weeks, and that is acceptable provided the rank is consistent. A slow but predictable seller is a perfectly good wholesale line, because you are buying it repeatedly rather than once.

A worked example

The product that passes shows: Amazon not on the listing, a consistent and good sales rank, a Buy Box price above the 15 floor, and only two competing offers.

The detail worth studying is the competition. The lowest price on the listing is a merchant-fulfilled offer at £24.30, but with delivery it comes to nearly £30, and the Buy Box is held by an FBA offer. When a merchant-fulfilled price plus delivery is close to the Buy Box price, the FBA seller normally wins the Buy Box. So the real competitor count is two, not the full offer count.

The final checks are on the Data tab: rank consistent over 90 days, Amazon out of stock 100 percent of the last 90 days, and an average Buy Box price near £40.

Record it

Every product that passes goes in a spreadsheet: link, product name, ASIN, average price, average sales rank, the rank percentage band it falls into, and a supplier column left empty for now.

That list is the input to the supplier search. You work through it product by product, searching the name plus "wholesale", using the methods from the previous lesson.

Mining a competitor's storefront

The second favourite method, with a specific rule for reading it.

Open a listing's offers and look at the sellers. Judge their size by review count over the last 30 days:

Under 100 reviews. Small. Almost certainly an arbitrage seller. Useful for arbitrage ideas, not for wholesale.
100 to 1,000 reviews. The target range for wholesale, with 300 to 600 described as the sweet spot. Large enough to be buying wholesale, small enough that their sources are reachable.
Over 1,000 reviews. A large operation, probably with exclusive arrangements you cannot replicate.

Click into a seller in that middle band and open their storefront. Everything they sell is a product somebody is already buying wholesale profitably.

The trap in storefront links, and how to avoid it

A specific technical point that will otherwise cost you real money.

When you open a product from a seller's storefront, the URL carries that seller's ID, and the page shows you only their offer. All the competition information is hidden. A product can look like it has one seller when it has forty-four.

The fix: edit the URL and delete everything after the ASIN, then reload. You now see every seller on the listing and can judge it properly.

In the example, doing this reveals 44 other sellers and that Amazon is on the listing, so the product is dropped immediately. Without the fix it would have looked like an uncontested find.

Two more lists worth keeping

Brands. When you find a brand with a good range, record the name. That list becomes your approach list for the Contacting Brands lesson.
Competitors. When a seller's storefront produces several good candidates, record their storefront URL, their seller ID (the value after me= in the URL), their name, their review count and the date you last checked. A productive competitor stays productive, and revisiting them later takes minutes.

What to expect from this

It is slow the first few times and becomes much faster with practice. It is also an ideal task to delegate, because it follows clear rules and needs no judgement beyond the criteria you have written down.

And every session produces two outputs rather than one: a list of products to find wholesalers for, and a list of products to find cheaper retailers for. The same work feeds both sides of the business.