On Amazon, the expensive mistake happens before your first sale. You pay for a few hundred units, ship them into a fulfillment center, and only then learn whether shoppers want them at a price that leaves a profit. If not, you pay storage on stock that sits.
Amazon product research is the work you do before that purchase order. This guide gives you a six-part framework, the numbers to check in each part, a pass mark, the red flags that should stop you and a worked example. A spreadsheet is enough, or pull most of the data for a niche into one report with our free Amazon product research tool.
The six-part research framework at a glance
Work through the parts in order. Each can end the research, and ending it at part one costs an afternoon, not a shipment. The pass marks are our working targets for a first product, not Amazon rules.
| Part | Question it answers | Data to check | Pass mark |
|---|---|---|---|
| 1. Demand | Is anyone searching, and is that growing? | Niche search volume, growth on last year, seasonality | Steady or rising searches, not one short spike |
| 2. Product DNA | What exactly do buyers choose? | Formats, median price and spread, complaints in reviews | A format you can make well, priced inside the main band |
| 3. Competition | Can a new listing reach page one? | Review counts, listing age, top brand's share, number of brands | Some newer listings selling, no brand with half the sales |
| 4. Keyword evaluation | Which searches can you win first? | Volume, growth, conversion, click share, page one per term | Five or more specific terms the leaders do not own |
| 5. Unit economics | Does each sale leave money after Amazon? | Landed cost, referral and FBA fees, storage, ads per unit | Net margin of 20% or more after ads |
| 6. Conclusion and launch plan | Go, test or walk away? | All of the above, order size, cash at risk | A written verdict, an order size and a keyword list |
One research session, in order:
- Write the niche the way a shopper types it, such as "bamboo drawer organizer".
- Check demand and its trend. Stop if searches are small or falling.
- Note formats, prices, review counts and age for the top 20 listings.
- Pull the niche's keywords and mark the ones you could rank for.
- Get landed-cost quotes from two suppliers and work out the margin.
- Write a go, test or avoid verdict before you place any order.
Sell on Etsy too? The Etsy product research tool builds the same report for Etsy phrases, and our guide to finding a profitable Etsy niche covers the rest.
Part 1: Demand
Research a niche, not a single keyword, because shoppers looking for the same product search with dozens of phrases. Amazon's Product Opportunity Explorer, under Growth in Seller Central, groups search terms into niches this way. Our report matches your phrase to the closest niche Amazon tracks and shows its name, so you can confirm it is the market you meant.
- Search volume. Enough that a small share of it would sell through your first order within a few months.
- Growth. Compare with a year earlier. Flat or rising passes.
- Seasonality. A one-season niche can work, but you then order, ship and stock on its calendar.
- Conversion. How often searches end in a purchase. A low rate can mean current products disappoint, which is a warning or an opening.
Red flags: a spike from one viral video, demand built on a brand or character name you cannot use, and a full year of decline.
Part 2: Product DNA
Product DNA is what the winning products share: format, size, material, pack count and price. A $12 single item leaves far less room for fees and ads than a $45 set with equal demand.
- Formats. Count the main product types among the top listings. If most sales go to one format, that is what buyers expect.
- Price bands. Build your margin from the median price and the band where most listings cluster, not from the price you hope to charge.
- Complaints. Read the leaders' one- to three-star reviews. A complaint that repeats, such as a lid that leaks, is an improvement you can sell on. Many good Amazon product ideas start there.
Red flags: prices sliding as sellers undercut each other, products that need approval, safety testing or certificates you cannot get, and items bulky or heavy for their price, since fulfillment fees rise with size and weight.
Part 3: Competition
Demand says a market exists. Competition says whether a listing with zero reviews can take a piece of it. Study page one for the niche's main searches.
- Review counts. Use the median, not the top seller. Thousands of reviews across page one means a new listing starts far behind.
- Age of the top sellers. Listings under a year old that already sell well show the door is open.
- Brand concentration. When one brand takes half the sales or more, you face its ad budget on every search. Many brands with modest shares point to an open market.
Our report's competition section shows the median review count, the mix of new, mid-life and established products, the top brand's share of sales and the number of brands. Like the product and fee sections, it comes from the niche's top 30 products, which the report reads live from Amazon the first time anyone researches that niche. Until then the report says so instead of showing zeros; check page one by hand or run the niche again a week later.
Red flags: one brand with most of the sales, Amazon's own brands on top and a median review count in the thousands.
Part 4: Keyword evaluation
A new listing rarely wins a broad head term first. It wins specific searches, earns reviews there and climbs. Those first targets shape your title, bullets and launch ads, so check each term for:
- Search volume, so the term earns a place in your title.
- Growth, so you are not building on a term shoppers are leaving.
- Conversion rate, how often a search for it ends in a purchase.
- Click share, as Amazon reports it for the niche. A high share marks a term that sends shoppers to products like yours.
Then search each shortlisted term on Amazon: if page one is all products with thousands of reviews, save it for later. The Amazon keyword research tool lists every search term Amazon tracks for a niche with its volume, growth, click share, sales and conversion, and marks golden keywords: terms with 500 or more searches that also show sales and conversions. Start your shortlist there.
Red flags: nearly all the volume on one or two broad terms, or the best terms contain brand names you cannot use.
Part 5: Unit economics
This part decides whether the business works. Work per unit:
- Landed cost: factory price, freight, duties, prep and labels, and shipping into Amazon's warehouses, divided by units ordered.
- Referral fee: a percentage of the total sale price or a per-item minimum, whichever is greater. At the time of writing (2026), most categories charge 15% with a $0.30 minimum; check Amazon's selling fees page for your category.
- FBA fulfillment fee: per unit sold, based on the product's size, weight and price. Just over a size tier boundary means paying more on every sale.
- Storage: monthly, on the space your stock takes up, with higher rates in the holiday months and a surcharge on units stored longer than 181 days.
- Advertising: expected ad spend divided by units sold, highest at launch.
- Account: the Professional selling plan is $39.99 a month at the time of writing.
Net margin is profit per unit divided by the price. ROI is profit per unit divided by landed cost: the return on the cash you tie up in stock. Put your price and costs into the Amazon FBA calculator to see profit per unit, margin, ROI and break-even price, free in your browser, then confirm the fees in Amazon's own Revenue Calculator before you order. Our research report adds a suggested price for the niche with its real median FBA and referral fees, read from its top products.
Red flags: a margin that only works with no ad spend, a price so low that fixed fees take most of it, and an ROI too thin to fund a reorder.
Part 6: Conclusion and launch plan
Write the verdict down. A decision you have to explain is harder to talk yourself into.
- Go: every part passes. Order enough stock to stay in stock through launch.
- Test: one number is thin, often the launch margin. Order a small batch and set the sales you need before reordering.
- Avoid: demand is falling, a few brands own the market or the margin fails after ads.
A go or test plan names the format and the improvement you sell on, the order size, five to eight target keywords, a daily ad budget and a review date. Our Amazon product research report ends with a draft of this: a verdict with strengths, risks, ways to stand out, launch steps and target keywords, drawn from the report's data. Treat it as a second opinion.
Worked example: bamboo drawer organizers
An example with made-up but realistic numbers, not data on a real niche.
- Demand: about 48,000 searches in the past 30 days, up 12% on a year earlier.
- Product DNA: expandable trays take most sales and fixed box sets take some. The median price is $29.99, with most listings between $24 and $36. Low-star reviews keep mentioning trays that slide.
- Competition: the median page-one product has about 1,100 reviews, but four listings under a year old already sell well. The top brand takes 22% of sales among 14 brands.
- Keywords: page one for "bamboo drawer organizer" is all products with thousands of reviews. "expandable bamboo drawer organizer" and "bamboo utensil drawer organizer" are golden and show newer listings on page one.
Per unit at a $29.99 price:
- Landed cost: $8.40 (product $5.60, freight and duty $1.90, prep and labels $0.30, inbound shipping $0.60)
- Referral fee at 15%: $4.50
- FBA fee, assumed for this size and weight: $6.80
- Storage, averaged per unit: $0.35
- Launch ads: $4.00
- Profit: $5.94, a 19.8% net margin and a 70.7% ROI
If ad cost falls to $2.50 a unit after launch, profit rises to $7.44, a 24.8% margin and an 88.6% ROI. The verdict is test: demand is growing and specific searches are open, but the launch margin sits just under 20%. The plan: an expandable tray with non-slip feet for the top complaint, 300 units, a title built on the two golden terms and a review after eight weeks before any reorder.
FAQ
What is Amazon product research?
It is checking a product idea against Amazon data before you buy inventory: demand, product formats, competition, keywords and the profit left after fees and ads, ending in a go, test or avoid decision.
How do I find a niche on Amazon?
Start with products you understand and could source, then judge each as a niche: steady or growing searches, newer listings selling on page one, no dominant brand and a margin after fees. Our product research tool works as an Amazon niche finder when you run several ideas and compare reports; each report's growing keywords point to narrower niches to try next.
Is there a free Amazon product research tool?
Yes. Our Amazon product research tool is free. Without an account you get one report a day showing the demand, the top five golden keywords and the verdict headline. A free account opens the full six-part report three times a day.
What profit margin should an Amazon FBA product have?
Amazon sets no target. We use a 20% net margin after referral, FBA, storage and ad costs as the floor for a first product, because launch ads and returns eat thin margins fast.
