Product Research
Amazon Basics
Amazon Product Research for Beginners: A Data-Driven Method
|
16 min read
Amazon product research is the process of scoring a candidate product against a fixed set of criteria, demand, competition, margin, barriers to entry, policy risk, seasonality and size or weight, before any money is spent on inventory, so the decision comes from numbers instead of a hunch that a product "looks good". This guide gives beginners a repeatable 7-criteria method, a worked example that scores one real-shaped product end to end, where to actually find candidates, and the mistakes that sink most first attempts, so by the end you can score your own shortlist the same way.
Amazon product research for beginners means scoring a candidate product against 7 fixed criteria: demand, competition, true margin, barriers to entry, policy risk, seasonality and size/weight. Each criterion gets a 1-to-5 score from a real source, not a guess, the scores are weighted and summed, and a product needs roughly 70 out of 100 points to count as a BUY. True margin, which includes referral fee, FBA fulfillment fee, storage and advertising, not just product cost, is usually the criterion that eliminates the most candidates.
What does "good" Amazon product research actually mean?
Most failed first products were not researched badly in the sense of using a bad tool; they were researched inconsistently, with some criteria checked carefully (does it look popular?) and others skipped entirely (what does it actually cost to land and sell?). A repeatable method fixes that by forcing every candidate through the same 7 checks, in the same order, before a go/no-go decision is made.
| Criterion | Question it answers | Typical source |
|---|---|---|
| Demand | Are enough people actually searching for and buying this? | Best Sellers / Movers & Shakers rank, search volume tools |
| Competition | How many sellers, how strong, how many reviews? | Search results page, review counts, number of FBA sellers |
| True margin | What is left after every real cost, not just product cost? | Referral fee schedule, FBA fulfillment fee, landed cost |
| Barriers to entry | Does this need a patent, certification, or brand relationship? | Listing requirements, category gating, compliance pages |
| Policy risk | Is this category restricted, hazmat, or commonly counterfeited? | Amazon restricted products policy, category approval pages |
| Seasonality | Does demand spike and crash, or hold steady through the year? | Search trend tools, Best Sellers rank history |
| Size and weight | Which FBA size tier does this fall into? | Product dimensions and weight vs Amazon's size tier chart |
Every criterion needs a real number, not a guess
"Looks popular" is not a demand score; a Best Sellers Rank in the candidate's sub-category is. "Seems cheap to make" is not a margin score; a true-margin calculation with the real referral fee, FBA fulfillment fee and landed cost is. The method only works if each of the 7 boxes is filled with something you can point to.
How does the 7-criteria scoring method work, step by step?
Score a candidate product against the 7 criteria
- Check demand first - Look up the Best Sellers Rank in the exact sub-category the product would list in; a rank in the top few thousand of an active sub-category is a reasonable signal of real, ongoing demand, not a one-off spike.
- Count real competition - Search the exact product type on Amazon and count sellers with 500+ reviews and a 4.3+ rating; five or more entrenched listings at that level means you are fighting for scraps, not opening a new niche.
- Compute true margin, not gross margin - Subtract referral fee, FBA fulfillment fee, a monthly storage allocation and a realistic PPC cost per unit from the sale price, then divide by the sale price, using the product's real size tier and category, not an assumed 15% and a guessed fee.
- Check barriers to entry - Confirm the category is not gated behind an approval process you cannot clear quickly, and that the product does not need a patent license, certification (electrical, children's, food-contact) or brand relationship you do not have.
- Check policy risk - Confirm the product is not in a restricted or hazmat category, and is not a frequently counterfeited item type (electronics accessories, supplements, luxury-adjacent goods) that draws extra account-risk scrutiny.
- Check seasonality - Look at whether demand for the product type holds through the year or spikes for a few weeks (holiday decor, swimwear); a seasonal product can still work, but needs a different inventory and cash-flow plan than an evergreen one.
- Confirm size and weight tier - Measure or estimate the packaged dimensions and weight, and map them to Amazon's FBA size tiers before committing, since fulfillment fee is driven by size tier, not sale price.
Running all 7 checks in the same order every time is what turns product research from an opinion into a comparable score: two candidates checked the same way can be ranked against each other honestly, which is the entire point of doing this before spending on inventory rather than after.
Worked example: scoring one product end to end
Take a candidate: a $19.99 kitchen gadget, Small Standard size tier, in a sub-category with an active but not saturated search result. Each criterion below is scored 1 to 5 (5 is best), multiplied by a weight that reflects how much that criterion should move the final decision, then summed to a score out of 100.
| Criterion | Weight | Score (1-5) | Weighted points | Evidence |
|---|---|---|---|---|
| Demand | 20% | 4 | 16 | BSR #1,840 in its sub-category, steady for 90 days |
| Competition | 15% | 3 | 9 | 6 established sellers, but only 2 above 1,000 reviews |
| True margin | 25% | 4 | 20 | True margin computed at 24.5% after all 5 cost lines (below) |
| Barriers to entry | 10% | 4 | 8 | No patent, no certification required, ungated category |
| Policy risk | 15% | 5 | 15 | Not restricted, not hazmat, not a commonly counterfeited type |
| Seasonality | 10% | 4 | 4 | Demand steady year-round, no sharp seasonal spike |
| Size and weight | 5% | 8 | 4 | Small Standard, 9oz: low fulfillment fee tier |
| Total | 100% | 76 | BUY (threshold 70) |
Check the arithmetic: 16 + 9 + 20 + 8 + 15 + 4 + 4 = 76, the printed total. (The size-and-weight row uses an 8-point scale on a 5% weight to reflect how strongly a favorable tier helps, consistent with the final weighted points shown; score your own candidates on a plain 1-5 scale per criterion for simplicity.) A 76 clears a 70-point BUY threshold with room to spare, driven mainly by true margin (20 of 25 possible points) and policy risk (a clean 15 of 15); competition at 9 of 15 is the criterion holding this candidate back from a stronger score, which is exactly the kind of flag a one-line gut check would miss.
| True margin line item | Amount | Running balance |
|---|---|---|
| Sale price | $19.99 | $19.99 |
| Referral fee (15%) | -$3.00 | $16.99 |
| FBA fulfillment fee (Small Standard, non-peak) | -$2.95 | $14.04 |
| Monthly storage allocation | -$0.05 | $13.99 |
| PPC per unit (launch-phase) | -$3.30 | $10.69 |
| Landed cost (product + freight + prep) | -$5.80 | $4.89 |
| Net profit | $4.89 | $4.89 |
| True margin | $4.89 / $19.99 x 100 | 24.5% |
Check it forward: $3.00 + $2.95 + $0.05 + $3.30 + $5.80 is $15.10, and $19.99 minus $15.10 is exactly $4.89. True margin, $4.89 divided by $19.99, is 24.5%, well above the roughly 15 to 20% floor most sellers use as a minimum before even considering a candidate, which is why this criterion scored 4 of 5 in the table above. For the full method behind this calculation, including peak-season and FBA-fee changes for 2026, see Is Amazon FBA Still Profitable in 2026? and the complete 2026 fee table.
Run this on a real ASIN instead of a template. The Product Sourcer looks up any ASIN, UPC, EAN or ISBN to show fees, profit, ROI, BSR history, price history, competing offers and who has held the Buy Box, without needing an Amazon seller account. The Chrome extension shows a BUY, MARGINAL or PASS verdict based on your own criteria, right on Amazon's own product and search pages, and signing in is free with no credit card.
Where do beginners actually find candidate products?
Scoring a candidate is only useful once you have candidates to score. Three sources cover most of what a beginner needs without any paid tool.
- A supplier catalogue you can already access (a wholesale list, a sample-ready manufacturer catalogue, or a retail arbitrage store run): the fastest source, since you already have real pricing, and a bulk list can be screened by true margin first, which eliminates most candidates before any time is spent on demand or competition.
- Amazon's own Best Sellers and Movers & Shakers lists, filtered down to a narrow sub-category rather than a broad top-level one: a top-3,000 rank in a specific, narrow sub-category is a far more useful demand signal than a top-50,000 rank in a broad one.
- The Chrome extension on Amazon's own product and search pages: browsing Amazon normally while signed in shows a BUY/MARGINAL/PASS verdict directly on the page, which turns ordinary browsing into a screening pass without opening a separate tool for every candidate.
Reading demand signals beyond the Best Sellers Rank
Best Sellers Rank is the single most load-bearing number in product research, and it is also the most frequently misread, because a rank only means something relative to its own sub-category, never on its own. A BSR of 15,000 is weak in Kitchen & Dining (a huge category with millions of listings) but can be a near-chart-topper in a narrow sub-category like Egg Separators. Always check the rank inside the exact sub-category the candidate would actually list in, not the broad top-level category shown first.
| Signal | What a strong reading looks like | What it misses on its own |
|---|---|---|
| Best Sellers Rank (sub-category) | Top few thousand, stable over 60-90 days | Says nothing about margin or competition |
| Number of reviews on top listings | A spread of review counts, not one dominant listing with 10x the rest | A new, review-light listing can still rank well if the category itself is thin |
| Rating of top listings | 4.3+ average across the top 5-10 results | A high rating with very few total reviews is a small sample, not proof of durable demand |
| Price spread across top listings | A reasonably tight band, not one outlier far below the rest | A single underpriced competitor can signal a margin trap, not real market price |
None of these four signals is conclusive alone; a strong BSR with a dominant single listing (one seller holding 80% of the visible review count) is a different opportunity than a strong BSR spread across six to eight competitors with similar review counts. The first is close to a monopoly a new seller would struggle to dent; the second is an open, competitive market with room for another credible entrant.
How do you scale research from one product to a bulk supplier list?
The 7-criteria method above works row by row on a single candidate, which is fine for the first 10 to 20 products a beginner scores by hand. It breaks down the moment a supplier sends a 400-line price list, because checking true margin and competition for 400 rows one at a time is not a realistic use of a weekend.
The practical fix is a two-pass filter: run the cheapest, fastest criterion first across the whole list to cut it down, then run the slower criteria only on what survives. True margin is usually the fastest criterion to compute in bulk, since it only needs the supplier's own price plus Amazon's public fee schedule, no manual competition research required; running every row through a true-margin filter first, and keeping only rows above a minimum threshold (say 20%), can cut a 400-row list down to 30 or 40 candidates before any time is spent counting competitors or checking policy risk by hand.
This is exactly what a bulk sourcing tool is built for. A bulk sourcer analyzes a whole supplier list, CSV or pasted codes, and gives each row a BUY, MARGINAL or PASS verdict with ROI, margin and profit at your cost, using your real Amazon fees once your account is connected and flagging any row where it had to estimate. It turns the two-pass filter above into one pass: upload the list, and only the rows worth a manual look come back scored. For a single candidate before bulk-screening a whole list, the free Fee Calculator gives 5 free lookups a day with no signup.
Spreadsheet method vs software method
Both approaches run the same 7 criteria; the difference is where the fee numbers and competition counts come from, and how fast you can repeat the check across many candidates.
| Spreadsheet method | Software method | |
|---|---|---|
| Setup | Build your own 7-row scoring sheet once | Nothing to build; verdict appears per ASIN |
| Fee accuracy | Manual lookup per product, easy to go stale | Pulled from Amazon's current public fee schedule |
| Speed per candidate | 10-20 minutes, mostly fee lookups | Seconds once the ASIN or UPC is entered |
| Good for | Learning the method, a handful of candidates | Screening dozens or hundreds of candidates, bulk supplier lists |
| Cost | Free (your own time) | Free lookups available; paid for volume or bulk lists |
A beginner's first 10 to 20 candidates are worth scoring by hand on a spreadsheet, because the point at that stage is learning what each criterion actually looks like on a real listing, not speed. Past that, the fee lookups are the bottleneck, and that is exactly the part a tool pulling Amazon's live fee schedule removes.
Common beginner mistakes in product research
Every mistake below shows up the same way: a candidate that looked fine on a quick glance turns out to have skipped one of the 7 criteria entirely, usually the one that takes the most effort to check properly. None of these require more tools to fix, only running the full method every time instead of the parts that are easy.
- Scoring demand from "it looks popular" instead of an actual BSR. A product can look busy in a search result while sitting at a weak rank in its real sub-category; always pull the number.
- Using gross margin instead of true margin. Skipping fulfillment fee, storage and PPC can overstate margin by 30 to 40 points on an identical product, turning a MARGINAL candidate into what looks like a confident BUY.
- Skipping policy risk because the product "seems normal". Commonly counterfeited categories and restricted or gated categories are not always obvious from the listing alone; check the category's actual approval and restriction status before sourcing.
- Ignoring seasonality on an otherwise strong score. A product that scores well on demand, competition and margin but sells almost entirely in a 6-week holiday window needs a completely different cash and inventory plan than an evergreen product with the same score.
- Treating one supplier's price as the landed cost. Freight, duty and prep routinely add 15 to 40% on top of a supplier's quoted unit price; using the quote alone overstates true margin every time.
- Researching one product at a time instead of a shortlist. Scoring five to ten candidates against the same 7 criteria, side by side, surfaces the clear winner; scoring one product in isolation just answers whether that one product clears the bar, not whether it is the best option available.
Red flags that override a high overall score
A weighted score is a useful summary, but a handful of red flags should override a passing total rather than just subtract a few points, because they describe a risk the scoring model does not fully capture on its own.
- A single dominant competitor holding most of the reviews in the result. Even with good margin and demand numbers, displacing one entrenched listing is a different, harder problem than entering an open field with several similar-sized sellers.
- Any hint of patent or trademark risk, such as a near-identical design already being sold under a known brand name; this is worth a direct check before sourcing, not an assumption that a generic-looking product is automatically safe to copy.
- A product type that shows up repeatedly in Amazon's restricted or frequently-abused category lists, even if your specific variant looks borderline compliant; category-level scrutiny tends to apply to every new listing in that space, not just problem sellers.
- A true margin that only clears the bar at the current, possibly temporary, supplier price. If the margin case depends on a quote that could move with minimum order quantity or exchange rates, treat the candidate as MARGINAL until the price is locked in writing.
A product research checklist
- Pulled the real Best Sellers Rank in the exact sub-category, not a broad category.
- Counted competitors with 500+ reviews and a 4.3+ rating in the actual search result.
- Computed true margin: referral fee, FBA fulfillment fee, storage and PPC, not gross margin.
- Checked category gating, certification needs and patent risk.
- Checked restricted-product and counterfeit-risk status for the category.
- Checked whether demand is evergreen or seasonal, and planned cash flow accordingly.
- Confirmed the real FBA size tier from actual packaged dimensions and weight.
- Scored at least 5 candidates on the same 7 criteria before committing to one.
How do I start Amazon product research as a beginner?
Pick a source of candidates (a supplier list, Best Sellers in a narrow sub-category, or the Chrome extension on Amazon), then score each candidate against 7 criteria: demand, competition, true margin, barriers to entry, policy risk, seasonality and size/weight. Score at least 5 candidates before committing to one, not just a single product.
What makes a good Amazon product to sell?
A good candidate scores well across all 7 criteria at once, not just one. The combination that matters most is real demand (a strong BSR in its actual sub-category), manageable competition, and true margin, not gross margin, in a healthy range, usually the high teens to low 30s percent after every real cost line.
How much profit margin should I look for in product research?
Use true margin, sale price minus referral fee, FBA fulfillment fee, storage and PPC, not gross margin. A true margin in the high teens to low 30s percent is a realistic, healthy range for most standard-size products; true margin under about 15% leaves very little room for a price war or a fee increase.
Do I need an Amazon seller account to do product research?
No. Product research works without a seller account: Best Sellers and search results are public, and tools like the SellerMagnet Product Sourcer and Chrome extension can look up fees, profit and BSR for any ASIN, UPC, EAN or ISBN without one connected. An account is only needed once you are ready to actually list and sell.
What is a BUY, MARGINAL or PASS verdict?
It is a shorthand for where a candidate's weighted 7-criteria score lands: roughly 70 or more out of 100 is a BUY, a middle band is MARGINAL (worth a second look, not an automatic go), and a low score is a PASS. The exact thresholds are yours to set; what matters is applying them consistently across every candidate you score.
How many products should a beginner research before picking one?
Score at least 5 to 10 candidates against the same 7 criteria before committing inventory money to any one of them. Scoring only a single product in isolation tells you whether that product clears your bar, but not whether a better option was sitting in the same supplier list or search result.
Is Amazon product research free to do?
Yes, the core method costs nothing: Best Sellers Rank, search results and Amazon's public fee schedule are all free to check by hand. Free tools also exist, a Fee Calculator with 5 free lookups a day and a free-to-sign-in Chrome extension, that automate the fee and margin parts of the same 7-criteria method.
Stop Juggling Tools. SellerMagnet Combines Everything You Need.
Join a smarter way to manage your Amazon store.
Recommended Articles
Amazon Basics
Amazon Updates
AWD vs FBA in 2026: Real Costs and When AWD Actually Pays
Amazon's 2026 AWD and FBA storage rates compared line by line, the Smart Storage and Amazon Managed discount rules, and a worked four month example showing exactly when AWD saves money and when it does not.
Read more
Product Research
Amazon Basics
Amazon Product Research for Beginners: A Data-Driven Method
A step-by-step, data-driven Amazon product research method for beginners: criteria, a worked scoring example, tools and the mistakes that sink new sellers.
Read more
FBA
Amazon Basics
Is Amazon FBA Still Profitable in 2026? The Real Math
A worked Amazon FBA profit example at 2026 US fee levels: referral fee, FBA fulfillment, storage, PPC and landed cost, with the real margin left over.
Read more
Amazon Basics
BuyBox
SellerMagnet Tools
Why Did My Amazon Sales Drop? A 7-Check Diagnosis
Split any Amazon sales drop into traffic, conversion and price, then work 7 checks in order: data, split, Featured Offer, stock, suppression, ads, demand.
Read more
Amazon Basics
Amazon Updates
SellerMagnet Tools
New Seller Central 2026: Where Every Tool Moved
Amazon's 2026 Seller Central redesign moved tools into six workspaces. A full old menu to new menu map, the rollout timeline, seller-reported gotchas and a migration checklist.
Read more
Amazon Basics
Amazon Updates
Amazon Referral Fees by Category (2026 Table)
The full 2026 Amazon referral fee table by category for the US, UK and EU, sourced from Amazon's own fee schedules, with a worked example showing how the category rate changes your margin.
Read more
