---
title: "The Science Behind Flapen's Niche Scores - Flapen"
canonical_url: "https://flapen.com/science"
last_updated: "2026-09-04T05:02:06.047Z"
locale: en
meta:
description: "How the Flapen Score works: ten weighted components, thresholds calibrated against 193,753 real niches, and a pass bar that only 4.8% of them clear."
"og:description": "How the Flapen Score works: ten weighted components, thresholds calibrated against 193,753 real niches, and a pass bar that only 4.8% of them clear."
"og:title": "The Science Behind Flapen's Niche Scores - Flapen"
---
Methodology
# The science behind our niche scores
A niche is a group of Amazon products competing for the same buyers. The Flapen Score grades every niche we track from 0 to 100 on one question: how good this market is for a new seller. Most research tools are built to find you something. Ours is built to throw things out. This page shows how the number is made, how to read it, and what to do with it.
[**Open the niche research**](https://flapen.com/research) **See how the score works**
- Niches scored
- **193,753**
- Pass rate
- **4.8%**
The data
## Where the data comes from
A score is only as good as its data. Our fleet collects Amazon niche data around the clock and rebuilds the public statistics twice an hour. Every threshold on this page was calibrated against those live niches. Not against opinion, and not against a spreadsheet someone built once and never checked.
193,753Niches scored
552,721US products tracked
1.40MUS weekly trend rows
*Verdict figures captured 2026-08-26. US catalog figures captured 2026-07-10. The fleet enriches continuously, so counts taken hours apart differ.*
The Flapen Score
## Ten checks, one number
Every check answers one part of the same question: is this a growing market where a new seller can win share at a profit? Ten weighted checks in five groups, demand, competition, entry, price, and risk, add up to one score out of 100. Across the brands we audited and ran, nine in ten failures would have been flagged by the score alone.
**Demand**
40 points
**Big Market**
15 points
How much shoppers spend on these products in a year. A big market means a winning product can earn a lot, which makes the risk worth taking.
**Trending**
15 points
How much searches have grown in the last three months. A growing market keeps creating openings for new sellers.
**High Conversion**
10 points
How often a search turns into a sale. When buyers buy fast, you spend less on ads to win each sale.
**Competition**
25 points
**Open Market**
15 points
The share of clicks going to the five most popular products. When no product dominates, it's easier to capture your share of the market.
**No Dominant Brand**
10 points
The share of clicks going to the five biggest brands. When shoppers aren't attached to one brand, a new brand can enter.
**Entry**
15 points
**Quality Gap**
10 points
The average star rating of the products already selling here. Read the negative reviews to see what to fix, and aim for a product 0.2 stars above the niche average.
**Low Review Barrier**
5 points
How many reviews the average product here has. That's roughly how many reviews your product will need before it sells at real volume.
**Price**
10 points
**Avg price**
5 points
What products here typically sell for. Between $15 and $100 there's usually enough room to make money after fees.
**Price range**
5 points
The spread between the cheapest and priciest products here. When shoppers already pay a range of prices, there's room to position your own, including a premium version at the top.
**Risk**
10 points
**Low Returns**
10 points
How often orders get returned. Every return costs you money, so a niche with rare returns protects your profit.
Price is the exception. Every other check climbs. Price peaks between $15 and $100 and falls away on both sides. Below the window, fees eat the margin. Above it, impulse buying fades and the cash tied up per unit grows. Both price checks cap at 95 instead of 100, because a good price is table stakes, not an edge.
Eight of the ten checks score on the same thresholds that grade the raw metric, so the badge next to a number and the points it earns never disagree.
A real niche, scored
## How to read the card
This is the real score card from the niche research, on live data. You get the same card for any niche we track. Read it in four steps.
1. 01
Start at the big number. That is the Flapen Score out of 100, with its verdict next to it: Pass, Maybe, or Fail.
2. 02
Read the line under it. It compares this niche with the others in its category, so you know whether the number is high or low for its neighborhood.
3. 03
Then look at the shape. Each axis of the radar is one of the ten checks. The dashed line behind it is the category average, so one glance shows where this niche is strong and where it trails.
4. 04
Hover any axis label to see what it measures, or any dot for the real number behind the points.
### **Kids Guitar**
**Maybe**
Shoppers spend $2.0M a year here. But demand is soft, at -7.6% this quarter, so Kids Guitar is a Maybe.
**Filters**[** No Dominant Brand**](https://flapen.com/research/niches?market=us&signal=no_dominant_brand "The top brand takes under 20% of clicks")
Flapen Score
**Maybe**
Top 1% of 437 Musical Instruments niches · +19 vs the category median of 44
Big MarketTrendingHigh ConversionOpen MarketLow ReturnsPrice rangeAvg priceNo Dominant BrandLow Review BarrierQuality Gap
This niche Category average
[**See the full report for this niche**](https://flapen.com/research/us/musical-instruments/kids-guitar)
### The dashed line is the average
Every category has its own normal. A big market in Pet Supplies is a small one in Toys and Games. So the radar always shows two shapes. The solid one is the niche you are looking at. The dashed one is the average of the niches in its category. Where the solid shape reaches past the dashed line, this niche is stronger than what is normal here. Where it stays inside, it trails. You can read a category you have never seen before in one look.
The Flapen Radar
## Two niches can score the same and look nothing alike
The Flapen Score tells you how good a niche is. The Flapen Radar tells you what kind of good: strong axes are the opportunity, weak axes are the plan. Weakness on the entry side can be worked through with patience and inventory. Weakness on the demand side cannot. No amount of seller skill creates buyers.
**A big market behind a wall**
Flapen Score **62**
Big MarketTrendingHigh ConversionOpen MarketLow ReturnsPrice rangeAvg priceNo Dominant BrandLow Review BarrierQuality Gap
_Demand carries this shape: a large, growing, converting market. The weak entry axes are the plan. Incumbents hold a review moat, so entering takes patience and inventory._
**An open door to a small room**
Flapen Score **62**
Big MarketTrendingHigh ConversionOpen MarketLow ReturnsPrice rangeAvg priceNo Dominant BrandLow Review BarrierQuality Gap
_Entry carries this shape: few reviews to catch up to and no dominant brand. But the demand axes say the market is thin, and no amount of operating skill fixes demand._
Illustrative shapes, drawn to land on the same score. Open any real niche in the research to see its own.
The verdict
## What to do with the number
Ten checks produce one number, and the number produces one of three verdicts. The bar is set against everything we track, so a Pass means roughly the top 5% of all niches: the typical niche scores 48, and even a niche that beats nine out of ten others sits at 61, still short of the line.
65+
**Pass**4.8% of all niches we track
Shortlist it. The market cleared the bar, so open the report and validate the rest: the products, the reviews, the price you can win at. From here the risk is your product and your plan, not the market.
50–64
**Maybe**39.7% of all niches we track
The common case. Read the radar for what holds it at Maybe: weakness on the entry side is a patience-and-inventory plan, weakness on the demand side is a reason to walk. The live example above is a Maybe for exactly this reason.
Below 50
**Fail**55.5% of all niches we track
Skip without guilt. The score exists to remove niches from your week, and a Fail is it working. Spend the attention on the short list instead.
A research tool that finds an opportunity everywhere you look is not measuring opportunity, it is measuring your willingness to keep scrolling. The value of this score is in the 184,000 niches it removes from your week.
Calibration
## Where we were wrong, and what this cannot tell you
We built the first version of these thresholds from an internal spreadsheet, then tested it against 15,742 real niches and found out how wrong it was. We would rather publish the corrections than have you find them, along with the things the score still cannot do.
### What we corrected - **Our size thresholds were off by an order of magnitude**
The sheet said a market needed $1M a year in sales to rate as merely acceptable. Across the niches we actually track that is the 92nd percentile, so the threshold graded roughly 90% of real niches as bad. We rebanded every size metric to the actual distribution. - **Our review thresholds described products that barely exist**
The sheet called 100 reviews an exceptional opportunity. The median product we track has 287 reviews, so 100 sits at the 32nd percentile, and the old rating bar would have graded 46% of real products as bad quality. - **We threw one-year growth out of the score completely**
It is the number every tool shows, and in the raw feed its semantics do not hold up: the median reads as +184% growth and near-zero-base artifacts run absurdly high. The score uses the 90-day window instead, which behaves like a real distribution.
### What it cannot tell you - **It does not price your product.**
Profit margin needs your cost of goods, your freight, and your ad spend. The score measures the market, not your P&L. - **Its thresholds are calibrated on US data and applied everywhere.**
The German market really is different, so German niches currently grade slightly harsh on size and slightly kind on returns. Per-marketplace thresholds are the next calibration, not a thing we have done. - **A passing score is a starting line.**
It says the market is worth entering. It does not say your product, your supplier, or your traffic plan will work. That part is operating, and operating is the harder half.
Results
## What the sellers say
[View story Grady's Pitching School**~~Mike Grady~~**
+30% year over year May was a great month. Up 30% over last year. Let's keep it rolling in June.](https://flapen.com/results/gradys-pitching-school)
[View story SnoreLessNow**~~Chris Garvey~~**
TACoS 10.8% → 9.9% Was burning hours on research that Flapen's AI does in minutes. Told me exactly what was worth pursuing and what wasn't. Wish I'd found it sooner.](https://flapen.com/results/snorelessnow)
[View story Tiny Tinker**~~Matt Dorband~~**
+41% YoY pace Flapen found baby product sub-niches I'd completely missed. Ran the profit calculator, numbers checked out, placed my first order with actual confidence.](https://flapen.com/results/tiny-tinker)
[View story Vora Bowl**~~Logan Leckie~~**
5.7% TACoS at scale Nice $3K day yesterday… new record?!](https://flapen.com/results/vora-bowl)
[View story Oral Pouch Solution**~~Brecht Powell~~**
Repeat buyers 7% → 14% Repeat customers doubled. That's when we knew it was working.](https://flapen.com/results/oral-pouch-solution)
[View story Silver Aid**~~Stephanie Hansen~~**
16.9% conversion rate Thank you for all of your hard work and continued support. We really appreciate everything you've done.](https://flapen.com/results/silver-aid)
[View story GrillX**~~Marko Tähtivaara~~**
ACoS 88% → 32% One tool instead of five. Research, forecast, optimize, done. Exactly what I needed.](https://flapen.com/results/grillx)
[View story Purefiz**~~John Mallia~~**
157 active subscriptions Didn't know where to start with research. Flapen connected the dots and by launch day the numbers already made sense.](https://flapen.com/results/purefiz)
[View story TuffTynz**~~Aubrey Laflamme~~**
9.5x creator-ads return Flapen found the product, launched it, and had us profitable in under 6 months. They now manage our 2 brands with 15+ SKUs — we're growing faster than we ever did alone.](https://flapen.com/results/tufftynz)
The method
## The score is one part of a bigger system
The method the number plugs into, and the story of how it was built.
[**~~The Flapen Story~~**
How a software engineer ended up launching 500+ Amazon brands, and why the whole method is public.](https://flapen.com/about)
[**~~The Flapen System~~**
The five steps between a product idea and a profitable brand, published end to end.](https://flapen.com/system)