Finding a Product Worth Your Cash

You are not hunting for a brilliant idea. You are hunting for proof that money already moves, plus a small opening where your version can be marginally better. Proven demand first, novelty never.

Start where the demand is provable

Type a category into Amazon's search bar and study page one like a competitor scout. What you want to see: multiple listings with substantial review counts, prices in a healthy band, and listings that clearly sell units week after week. Amazon's Best Sellers Rank, the BSR printed on every product page, tells you how fast an item sells relative to its category; lower is better, and a BSR chart over time, which Helium 10's free browser extension shows, reveals whether sales are steady, climbing, seasonal spikes, or a dead product coasting on old rank. A consistently low, steady BSR line beats a spiky one every time.

The paid tools, Jungle Scout and Helium 10, estimate monthly unit sales from BSR. Operators consistently treat one of the two as worth a month of subscription during research, then cancel; Jungle Scout's middle plan runs about $79 a month and Helium 10 runs from a free tier to $279 a month for its top plan billed annually. The free tier plus the free extensions can carry your first pass. Do not subscribe for a year on day one.

The quiet filters that do the real work

Product research courses love flashy criteria. The filters that save you are boring.

Price band: roughly $20 to $50. Below fifteen dollars, the fee stack from "What the Money Actually Looks Like" eats the margin; there is not enough room after a fifteen percent referral, a fulfillment fee, and ads. Above fifty or so, your first order's capital requirement blows past a lean budget, and returns hurt more.

Small, light, and sturdy. FBA fulfillment fees scale with size and weight, so a compact, durable product ships cheap and survives the warehouse. Nothing fragile, nothing oversized, no glass.

Simple. Your first product should have no electronics, no batteries, no liquids, no chargers, no moving parts you cannot inspect at the kitchen table. Complexity multiplies defect rates, and defect rates become returns and suspended listings.

Ungated. Categories like groceries, supplements, and top brands are gated, requiring approvals, invoices, or certifications. Search "Amazon gated categories" in Seller Central help before you fall in love with a product; the list lives there and shifts over time. Some gates are climbable; none are worth your first order. If the product pulling at you is edible, that pull points at a CPG food product business, where the regulation is the business rather than a gate around it.

Not contested by giants. If Amazon's own brand or a household name owns the top four results with thousands of reviews and lower prices than you can land, walk away. You are looking for pages where the top listings are ordinary sellers with five hundred to two thousand reviews, not fortresses. Know where that band sits: most current research guides draw the fair-fight line lower, under five hundred reviews, so treat two thousand as this course's permissive ceiling rather than a consensus.

Room to be different. This is the opening. Read the one-star and three-star reviews of the top ten sellers. Complaints are a product specification written by the market: handles that break, sizes that run small, instructions that confuse. Your version fixes two of them and says so in the listing.

Kill criteria, written down before you research

Bias will tell you your product is special. Write the kill list first and obey it: seasonal-only demand; a trademarked brand term anywhere near your intended name, check the USPTO search; top sellers priced below your achievable landed cost; fewer than a handful of sellers moving units, which means no proven demand; and any product where you cannot explain, in one sentence, why your version differs.

The shortlist discipline

Do not find one product. Find ten, run every one through the unit-economics calculator, and let the calculator pick. Your job in research is to generate candidates and kill them dispassionately. The calculator's job is to tell you which survivor deserves your two thousand dollars.