Sourcing With Arbitrage Hero

Manual sourcing works and it is slow. This lesson covers the automated version: setting the filters in Arbitrage Hero so the software compares hundreds of retailers against Amazon and hands you a shortlist.

What the software is doing

Exactly what you did by hand in the manual sourcing lesson, at scale. It holds retailer catalogues and Amazon data, matches products between them, works out the fees and the margin, and shows you the ones that pass your filters.

That is why the manual lesson comes first. The output is a list of candidates, not a list of instructions, and you still make the buying decision.

Choosing where to search

Two ways in:

• By category. Search across all supported stores within a category such as Kitchen. Best when you know what you want to sell.
• By store. Search one retailer's whole catalogue. Best when a shop is running a sale, or when you have a discount code for it.

The filters, and what to set them to

These are the starting values recommended in the lesson.

• Sales rank percentage. Defaults to 5 percent, meaning only products ranking in the top 5 percent of their category are returned. This is your demand filter. Tighten it to 2 percent for fewer and better-selling products; loosen it for more results and more risk.
• Minimum return on investment: 30 percent. A sensible floor while you are learning, because it leaves room for the price to move against you between ordering and selling.
• Minimum profit: 3 in your currency. This works with the return filter rather than duplicating it. A 40 percent return on a £4 item is 1.60, which does not cover the effort of handling a unit. The profit floor removes those.
• Exclude Amazon: yes. Leave this on. Competing against Amazon on its own listing generally means holding stock that does not move.
• Discount percentage. If the retailer has a code today, put it here. The results are then calculated against the discounted price, so products that fail the profit filter at list price start appearing. This is the setting from the Getting Ready lesson, and it is where the preparation pays off.

Matching score, the setting worth understanding

Defaults to 75. It controls how confident the software has to be that a retailer product and an Amazon listing are the same item.

• Higher means more accurate matches and fewer results.
• Lower means more results, some of which are wrong matches.

The advice in the lesson is to start at 75, then deliberately search at a lower score as well. The reasoning: some genuine matches are filtered out unfairly at a high score, and those are products nobody else is seeing, because everyone else leaves the default alone.

The cost is that you must verify the match yourself before buying. At a low score, some results are simply the wrong product.

Connecting your other tools

If you hold a SellerAmp SAS or BuyBotPro licence, enable the corresponding setting. The links in your results then carry the figures across, so opening a product on Amazon or on the retailer's site shows the extension already populated with the cost price and the calculation. It removes the retyping, and it means the Keepa chart and the sales estimate are in front of you immediately.

Reading the results

Results default to sorting by profit, high to low. Change it. This is the practical advice of the lesson: everybody who uses the default sees the same products in the same order and competes for the same deals.

Sort by sales rank, low to high, to see the fastest-selling items first. Try other orders. Search the same store at different matching scores. The deals nobody else has found are the ones sitting outside the default view.

Also ignore the extremes. A list sorted by profit puts very expensive items at the top, and a £2,000 product is not a sensible first purchase whatever the margin says.

Judging a candidate

The example in the lesson is a good one: a frying pan where Amazon appears on the listing and then leaves it repeatedly. The chart shows the price rising whenever Amazon is absent.

That pattern is worth recognising. Amazon coming and going is not automatically a rejection; it means your selling window is when Amazon is out of stock, and the price is better then. What you are judging is whether those windows are frequent and long enough for your stock to sell through.

The sales estimate is the other number to check. An item selling around 52 times a month has room for several sellers. One selling twice a month does not.

The most important habit in this lesson

List the item on Amazon before you buy it.

Not after it arrives. Before you pay the retailer. The reason is eligibility: as the gating lesson explained, restrictions are personal, so a product you can see and price may be one you are not permitted to sell. Discovering that after the stock is delivered leaves you holding goods you cannot list.

The check takes seconds. Scan the barcode from your screen with the Amazon Seller app, and it tells you immediately whether you can list the item or whether it requires approval. Do it every time, on every product, before every purchase.