How to Find Winning Products to Sell on Amazon A Screening Method, Not a Trend List

Amazon changed its non-peak FBA fulfillment rates on January 15, 2026. Standard-size products priced from $10 to $50 increased by $0.08 per unit on average, while standard-size products below $10 increased by $0.05 on average. Amazon also said it was introducing no new fee types in the 2026 US fee update.

Small numbers, but the architecture built since 2024 is what decides whether a product clears: inbound placement fees, the low-inventory-level fee, returns processing, aged inventory surcharges, and a fulfillment rate card now split by price band. Most product research guides stop at "find high demand, low competition." That advice was incomplete even before those fees existed. Below is the screening sequence we run, in the order that kills bad ideas fastest.

Start with Amazon's own data rather than a third-party estimate of it. Product Opportunity Explorer organizes demand into niches, which are clusters of search terms leading to substantially the same set of clicked products, and refreshes niche information weekly. Amazon is also explicit that an opportunity surfaced in the tool is not guaranteed to succeed.

Pull these for any niche you are considering:

• Search volume and search volume growth over the past 90 days against the prior 90
• Units sold in the niche over 90 days
• Average price across niche products
• Number of products in the niche, and how many receive 80% of clicks
• Recent launches into the niche, and how they have performed

Amazon's own guidance points at a specific gap worth hunting: products that receive most of the clicks in a niche but are frequently out of stock. That is demand nobody is currently serving.

A CMO warning signal at this gate:

Flat or declining search volume growth combined with rising product count. That is a niche filling up while demand stays still.

High competition is not automatically disqualifying. Concentrated competition usually is.

If a small group of ASINs absorbs most of the clicks and those listings also have established review profiles, treat the niche as concentrated. Before entering, identify a specific product, positioning, pricing, or audience gap that gives shoppers a reason to choose a new listing.

Two more checks worth running before you commit. First, compare New Products Launched against Successful Launches in the niche. Amazon counts a successful launch as one reaching annualized revenue of $50,000 or more, so a niche showing 40 launches and three successes over the same window is telling you what the odds look like for entrant number 41. Second, check the price band where clicks concentrate. If that band sits well below what your cost structure supports, the niche has already answered your pricing question.

This is where most product ideas die, and where most guides go quiet.

Fee	What Triggers It
Referral fee	Percentage of total sales price, with a $0.30 minimum
FBA fulfillment fee	Size tier, weight, and since January 2026 a price band
Fuel and logistics-related surcharge	3.5% of applicable FBA fulfillment fees beginning April 17, 2026
Inbound placement fee	Shipment split, size tier, weight; zero if you ship to enough locations
Monthly storage	Average daily cubic feet, higher in Q4
Aged inventory surcharge	Inventory held past the published age thresholds
Low-inventory-level fee	Historical days of supply under 28, now measured per FNSKU
Returns processing fee	Returns above your category's threshold
Refund administration fee	A share of the refund when a customer returns

Work a real example before sourcing.

A $29.99 product in a category with a 15% referral fee gives up about $4.50 to the referral fee alone. From there, model the actual FBA fulfillment fee using the product's dimensions, shipping weight, and applicable price band, then add the fuel and logistics-related surcharge, estimated inbound placement cost, storage, landed product cost, expected returns, and advertising. Use Amazon's Revenue Calculator and the fee and economics preview tools in Seller Central rather than a generic per-unit estimate.

Finishing that example.

Using illustrative fees for a one-pound standard-size product, the stack looks like this: $4.50 referral, $5.50 fulfillment, $0.19 surcharge, $0.30 inbound placement, $0.15 storage, and $0.60 set aside for returns. That is $11.24 of Amazon cost. Add $7.00 landed product cost and total cost per unit reaches $18.24, leaving $11.75 of contribution on a $29.99 price, or 39%. That clears the screening threshold with room for advertising, so the product moves to gate four.

Now change two inputs. Landed cost of $10.00 instead of $7.00, and a size tier that pushes fulfillment to $7.30. Total cost becomes $22.84 and contribution falls to $7.15, or 24%. Same price, same demand, same niche, and the product no longer clears. That is the entire value of running gate three before a purchase order rather than after the first shipment lands. Verify your own fee amounts in Amazon's Revenue Calculator; the figures above are illustrative.

As a CMO screening rule,

We generally want to see roughly 25% to 30% contribution margin after Amazon fees and product costs but before advertising before becoming comfortable with a launch. That is an operating threshold, not an Amazon benchmark. The appropriate margin depends on category, advertising requirements, return rates, working capital, and the brand's economics.

Two fee details worth designing around:

The low-inventory-level fee now measures at the FNSKU level.

A parent ASIN that looks healthy overall can still incur the fee on the one variation that dipped below 28 days of supply. Variation-heavy catalogs carry more of this exposure than single-SKU products. Our guide to preventing FBA stockouts covers the forecasting side.

Inbound placement is a design decision.

Shipping to more locations reduces or removes the fee. Whether that is cheaper than paying it depends on your freight setup, so price both before your first shipment.

Some products fail regardless of demand and margin. Check these before you get attached:

• Gated categories and brands. Some categories require approval, and some brands are restricted entirely. Check eligibility in Seller Central before sourcing.
• Hazmat, batteries, and meltables. Each carries extra handling requirements, seasonal restrictions, or documentation that adds cost and delay.
• Oversize and bulky. Fulfillment and placement fees scale hard with size. A product that works at standard size can be unviable one tier up.
• High return exposure. Check Product Opportunity Explorer and your category economics for return behavior. Returns can add processing costs and reverse-logistics losses that materially change the margin you modeled.
• Sub-$10 price points. Low Price FBA rates exist, but thin absolute margins leave little room for advertising.
• IP exposure. Check for relevant patents, trademarks, copyrighted material, and other intellectual-property risks before sourcing. A promising demand signal does not outweigh infringement risk.

Product Opportunity Explorer shows demand. It cannot tell you whether you can win.

Differentiation.

If your version is the same product with a different logo, you will compete on price forever. Read the negative reviews in the niche. The complaints that repeat across the top ASINs are your product brief.

Supply reality.

Minimum order quantities, tooling costs, production lead times, and freight all decide whether you can restock before you trip the low-inventory-level fee. A 90-day replenishment cycle changes which products are viable.

Review velocity.

New products compete against listings with years of reviews. Budget for the time it takes to build social proof, not just the launch.

Capital.

Inventory, advertising, and a reorder before the first batch sells through. Underfunding the reorder can stall an otherwise promising product just as demand begins to build.

Gate	Pass	Warning	Stop
Demand	Stable or growing demand supported by Amazon data	Seasonal or inconsistent	Declining demand with growing supply
Competition	Clicks distributed with an identifiable product gap	Moderate concentration	Highly concentrated with no clear differentiation
Margin	Clears your required contribution-margin threshold	Works only under optimistic assumptions	Negative or insufficient margin under realistic fees
Constraints	Straightforward eligibility, fulfillment, and compliance	Added documentation or cost	Restriction, IP, logistics, or compliance issue makes launch impractical

A product should not move to sourcing because it scored well overall. One hard failure can outweigh three strong gates. Document the reason for every pass, warning, or stop before approving a purchase order.

FBA New Selection Program.

Before launching an eligible new-to-FBA ASIN, check the current FBA New Selection benefits in Seller Central. Available incentives can materially affect early-launch economics, but eligibility, benefit periods, and applicable fee reductions can change, so include the current program benefits in your margin model rather than assuming standard FBA costs.

The listing, before the inventory lands.

Titles, images, and A+ Content should be finished before your first units are receivable. Our listing optimization checklist and guide to turning specs into benefits cover what that looks like. Set advertising targets your margin supports using our ACoS benchmarks.

CMO screening rule: run the gates in order and stop at the first failure.

We do not complete a full analysis on every candidate, because that is how research turns into months. Gates one and two take under an hour for a niche you already understand. Gate three is the real work, because it needs dimensions, weights, landed costs, and a fee lookup per candidate. Gate four is a checklist. Most ideas stop at gate three, and finding that out before a purchase order is the entire point.

What makes a product a "winning" product on Amazon?

One that clears all four gates. Demand you verified in Amazon's own data, competition that is not concentrated in a few entrenched ASINs, margin that survives the fee stack with room for advertising, and no disqualifying constraint. A product can have excellent demand and still be a bad idea if it is oversize with high return exposure.

Do I need Product Opportunity Explorer, or are third-party tools enough?

Third-party tools estimate what Amazon reports directly. Use Product Opportunity Explorer for niche demand and click distribution, and third-party tools for cross-checking and for research workflows Amazon does not support. Just know which numbers are measured and which are modeled.

How much margin should a product have before I commit?

There is no universal number. Model contribution margin after all fees and before advertising, then ask whether what is left funds a launch. Products that only work at zero ad spend rarely work at all.

How much capital do I need to launch a product?

Enough for the first order, the advertising behind it, and the reorder before the first order sells through. That third component is the one that sinks launches. Size it from your own model: units in the first order multiplied by landed cost, plus creative and sampling, plus a launch advertising budget your margin can carry, plus the reorder placed while the first batch is still selling. If the numbers only work when the reorder is funded by first-order revenue, the plan has no room for a delay at customs.

Should I pick a trending product or an evergreen one?

Trends reward speed and punish slow supply chains. If your lead time is 90 days, a trend will often peak before your second order lands. Evergreen demand suits most brands better, with trend products as a deliberate bet rather than a default.