Define the buying filter
Set the constraints before searching: target marketplace, product condition, category, price range, minimum margin, minimum ROI, acceptable size and weight, cash available and risk categories to avoid. A filter prevents an attractive but unsuitable product from taking over the research process.
Use permitted sources
Use retailer sites, supplier catalogues, approved feeds and properly licensed research tools. Record the source and the time checked. Do not scrape or republish restricted marketplace content, images or reviews outside the relevant terms.
Confirm the product identity
Match the source item to the Amazon listing using the EAN, ASIN, brand, model, pack size, variation, condition and included accessories. Amazon describes product identifiers and offer details as core listing information.[3] A title match alone does not confirm the product.
Check whether it can be sold
Review category, brand and product eligibility in Seller Central. Amazon says some categories require approval and some products cannot be sold by third-party sellers.[3] Check FBA restrictions before buying, not after the stock arrives.
Shortlist demand and competition
Review current price, price history, sales-rank context, offer count, Buy Box behaviour and Amazon presence. Treat estimates and historical charts as evidence to review rather than a forecast. Consider whether the product is seasonal or likely to fall in price.
Calculate the decision
Include buy price, delivery, preparation, inbound freight, referral fees, fulfilment, storage, returns, advertising, VAT, import charges and other costs. The FBA cost guide explains the main cost categories, and the profit calculator can test your assumptions.
Check risk before buying
Look for IP, authenticity, Hazmat, batteries, expiry, safety documentation, fragile packaging and condition restrictions. If the supplier evidence is weak or the product identity is uncertain, reject the lead even if the spreadsheet shows a strong margin.
Record and recheck
Save source URL, time checked, product identifier, prices, assumptions, expected margin, expected ROI, maximum buy price, quantity, risk flags and recheck date. A lead without a timestamp and assumptions cannot be reviewed properly later.
Use a controlled decision queue
Classify leads as new, needs review, approved to buy, rejected, no longer available or follow-up. The researcher prepares the evidence; the owner makes the cash-commitment decision. Keep the final buying decision under human control.