Most product research advice fails at the same point: it identifies what sells rather than what is worth selling. Those are different questions, and the gap between them is where new sellers lose money.
A product that sells 400 units a month at 4 percent net margin is a worse business than one selling 60 units at 28 percent. This guide is about finding the second kind.
The Screening Criteria That Actually Matter
Margin after every fee
Not sticker margin. Net margin after the referral fee, fulfillment cost, storage, and a realistic returns allowance for the category.
Target 20 percent net margin or better, or 30 percent ROI if you think in terms of capital returned. Anything thinner is fragile: a competitor undercutting you by 8 percent or a return rate 3 points above your estimate erases it.
Run every candidate through the FBA calculator and the profit calculator before it enters your shortlist.
Real sales velocity
Best Sellers Rank is the accessible proxy, with one caveat that matters enormously: BSR is only meaningful within its own category. A rank of 80,000 in a huge category can mean several sales a day; the same rank in a small category can mean one a week.
Learn the ranges for the two or three categories you intend to work in. A cross-category rank threshold is a way to be confidently wrong.
Two corroborating signals worth checking:
- Review accumulation rate. A listing gaining reviews steadily is making sales steadily.
- Seller persistence. If the same sellers have competed on a listing for six months, the economics work. If sellers cycle in and out every few weeks, they are discovering something you have not yet.
Competition depth
The workable range is roughly three to eight active sellers.
Fewer than three often means gating, brand exclusivity, or a product nobody wants. More than ten means the listing is already a price war, and price wars are won by whoever has the lowest cost basis — which, if you are new, is not you.
Check who currently holds the Buy Box and at what price using the Buy Box checker. If one seller has held it continuously, they likely own the brand and you are not going to displace them.
Price band
Between roughly 20 and 70 dollars is the sweet spot for most models.
Below 20, Amazon's fees consume too much of the sale and every return is disproportionately expensive. Above 70, purchase decisions slow down, return rates climb, and each unit ties up more of your capital.
Size and weight
Small and light is not a preference, it is an economic requirement. Fulfillment fees scale with dimensions, storage is charged by cubic foot, and oversize surcharges are brutal. Two products with identical retail prices can have completely different economics purely because of a box size.
Categories Worth Considering
Home and kitchen — enormous demand, endless product variety, mostly ungated, forgiving return rates.
Tools and home improvement — durable products, low return rates, buyers who know what they want. Regional clearance at home improvement retailers makes this a productive online arbitrage category.
Pet supplies — consumables mean repeat purchases and predictable demand.
Office and school — highly seasonal but genuinely predictable, which makes inventory planning easier than it looks.
Sports and outdoors — seasonal swings, but strong margins in the right window.
What to Avoid
Gated brands, before you have invoices. Check listing eligibility in your seller account before you buy anything, every time.
Hazmat and restricted items. Batteries, aerosols, flammables. The compliance overhead is not worth it while you are learning.
Fragile products. Damage in transit becomes returns, and returns are where thin margins die.
Heavily counterfeited categories. High-value electronics and designer goods attract counterfeit claims that are expensive to defend even when you are entirely in the right.
Anything where one seller owns the Buy Box permanently. They own the brand. You are not going to win.
Trend products with no history. By the time a trend is visible enough for you to source it, the price collapse is usually already scheduled.
Validating Before You Spend
Once a product clears screening, three checks before capital leaves your account:
Confirm you can list it. Eligibility is account-specific. Check in your own seller account, not a forum post.
Verify your supplier cost is real and current. A promotional price you screenshotted last week may not exist today. If you are sourcing online, this is not a one-time check — supplier prices drift, and a margin calculated against a stale cost is a fiction. This is exactly what supplier price monitoring is for.
Model the price war. Ask what happens if the price drops 15 percent. If you are underwater at that price, you have no room to compete, and the listing will get there eventually.
Start Narrow
The most common failure in product research is breadth. A hundred products sourced in a month, none of them understood, none of them monitored, and no idea which ones are actually making money.
Five to ten products you know well beats a hundred you cannot watch. You learn the seasonality, the competitor behaviour, the return rate, and the supplier's reliability — and those lessons transfer to the next fifty.
Then, when you scale, scale the monitoring at the same time. A catalogue of 200 products where costs move and competitors reprice hourly is not a manual job. Set floors from real costs, let a repricer work between them, and monitor your suppliers so a cost change or a stockout reaches your listings before it reaches your P&L.
Frequently Asked Questions
What is the most profitable thing to sell on Amazon?
Profitability comes from the gap between your cost and the market price, not from the category. In practice the most consistent margins come from products that are unglamorous, moderately priced between 20 and 70 dollars, small and light, and not dominated by a single brand.
How do I know if a product will sell on Amazon?
Use Best Sellers Rank as a velocity signal within its own category, check that several sellers have been competing on the listing for months rather than weeks, and look at review accumulation rate on the ASIN. A listing gaining reviews steadily is a listing making sales.
What products should I avoid selling on Amazon?
Avoid brand-gated categories until you have invoices, anything hazmat or requiring special handling, fragile items with high damage rates, heavily counterfeited electronics, and any listing where one seller holds the Buy Box permanently because they own the brand.
How many products should I start with?
Start with five to ten products you can actually monitor and understand. Breadth without process is the most common way new sellers lose money, because problems in a catalogue you cannot watch stay invisible until they are expensive.
Do I need a product research tool?
A research tool speeds up screening but does not replace fee arithmetic. Whatever tool you use, run every candidate through a real fee calculation including returns before committing capital.
Repricefy keeps every listing above a floor built from your actual landed cost, and tells you when a supplier's price moves that cost. Free for your first 25 products. Start free.