Browse the complete products index →
1,329,830 products scored · 1,509,012 examined
Products
Top categories
Automotive136,851Home & Kitchen121,687Office & Industrial108,740Toys, Games & Music72,718Patio, Lawn & Garden65,361Grocery & Gourmet Food54,060Pet Supplies49,828Electronics & Phones48,840Arts, Crafts & Sewing46,415Beauty & Personal Care44,442Health, Household & Baby29,325Tools & Home Improvement21,629Clothing, Shoes & Jewelry14,725Sports & Outdoors6,493
Top-ranked products
See all 1,509,012 →for Milwaukee 48-39-0521 44-7/8-Inch, 18 Teeth per in, Bi-Metal Band Saw Blade, 3PK
42oz Collapsible Silicone Steamer for Microwave(Grey) | Food-Grade Foldable Silicone Contanier for Cooking, BPA-Free & Heat Resistant Steam Pot, Dishwasher Safe
8-Piece Silicone Spatula Set, 600°F Heat Resistant, Food Grade Silicone Outer Layer, Earth Tone Dominated, Non Stick Utensils for Cooking & Baking
EW-DP ME2 Set (R1-6)
Beano's Classical White Pizza Saue With Basil, 32 Ounce Bottles
Cosco Kids™ MightyFit™ LX Convertible Car Seat, Canyon
Magnetic Power Strip with 6 Outlets & 4 USB Ports | USB-C PD 20W, 1800 Joules Surge Protector, Heavy Duty Extension Cord, Flat Plug Wall Mount for Garage, Workshop, 6FT, Black
WiFi Antenna RP-SMA 2.4GHz 5GHz 3dBi | for Backup Camera, Desktop PC, Motherboard, PCIe Adapter, Monitor
Animal Crossing: Happy Home Designer Amiibo Cards Pack - Series 2 (Nintendo 3DS/Wii U)
DERTE 4 Pcs Mattress Topper Straps Adjustable (78''x2'') | Buckles Memory Foam Mattress Mattress Packing for Storage and Moving Dark Grey
SAMSUNG 65-Inch Class QLED 4K Q70C Series Quantum HDR, Dual LED, Object Tracking Sound Lite, Q-Symphony, Motion Xcelerator Turbo+, Gaming Hub, Smart TV with Alexa Built-in (QN65Q70C, 2023 Model),Black
Tonies Anna Audio Toy Figurine from Disney's Frozen II
What goes into the score
| Reviewer normalization | Every review is measured against the baseline of the reviewer who wrote it, so a 4-star from someone who never gives 5s outranks a 5-star from someone who gives nothing else. This is the same protocol used for films, hotels, wines, cruises and books. |
|---|---|
| Reviewers who count | Anyone who rated at least two products — a single rating gives no baseline to measure against, so it is held back rather than counted. A reviewer whose ratings are all identical still counts: their dispersion is borrowed from the reviewer pool, and each of their reviews reads as "exactly par for me", which adds evidence without pushing the product up. That is 1,706,222 reviewers today, and it grows with every deeper pass over the reviews. |
| Global percentile | The headline: the average of a product's re-centered scores, ranked across every scored product. |
| Cohort percentile | The same average re-ranked inside the product's own category. |
| AI-adjusted percentile | The same average with thin samples pulled toward the corpus mean: a product keeps the share n/(n+53) of its measured distance from average, where n is its calibrated-reviewer count. A product measured by one reviewer keeps 1/54 of that distance; one measured by fifty keeps roughly half. |
| Pages we index | Products measured by two or more calibrated reviewers — 821,114 today. A product measured by exactly one still gets a page and still publishes all three percentiles; it is left out of the index because one reviewer is one reviewer. |
| Unscored listings | A listing whose reviews come only from one-off reviewers gets a page with no rating at all — no Rankquant percentile, and no borrowed star average standing in for one. Those pages are not indexed, and they gain a percentile when review coverage reaches them. |
| Reviews we read | Each page states how many reviews Rankquant collected and scored for that product, and how many of those came from calibrated reviewers. That is a sample size, not a rating. |
| Amazon’s own star average | Not published anywhere on this site. It is collected as a scrape-health diagnostic — an impossible value is the fastest way to spot a parser reading the wrong part of a page — and never reaches a page, a data file, a sort order, a filter or structured data. |
| Excluded listings | One rule, and it is about identifiers rather than ratings: ISBN-numbered listings are published media and belong to the Books catalog. Products are no longer withheld for a suspicious scraped rating average or rating count — those fields are never published, so a wrong value in one cannot mislead anyone, and withholding on it would let Amazon’s number decide our catalog. |
How the score works
- score1 — 0–100 percentile of the average re-centered score, across every scored product; the headline (top 5% = 95)
- Category cohort — the same figure re-ranked within its category
- AI-adjusted — the same average with thin samples pulled toward the corpus mean by n/(n+53)
- Reviews used — how many reviewer-normalized reviews stand behind it
- Reviews we read — the size of our own sample for that product, which is a count and not a rating
A cordless drill and a phone charger can share a 4.6-star average and still sit at very different cohort percentiles, because chargers as a category run hotter. Every scored product page shows both numbers side by side: the global percentile and the cohort one. That gap is the information the raw star average throws away.
What products share with the rest of the site, and what they don't: the normalization described on the methodology page— every review re-centered against the baseline of the person who wrote it — is exactly what runs here, and products rank the average of those scores, as wines, films and books do. What is different is coverage. Each listing has been read one review page deep so far, so most reviewers appear once and cannot supply a baseline; a product carries a percentile only once at least one of its reviewers rated something else too. Listings with no such reviewer stay published with Amazon's own figures and no Rankquant percentile, and move into the scored set as coverage deepens.
Frequently asked questions
Why not just trust the Amazon star average?+
Because it's compressed and category-blind. Most products with real volume sit between 4.3 and 4.7 stars, and a 4.6 in an easy category (phone cases, cables) is not the same as a 4.6 in a demanding one (laptops, mattresses). The percentile inside a category peer set restores the resolution the raw average loses.
How do you handle fake and incentivized reviews?+
Two mechanisms in the scoring reduce their effect. A reviewer who gives nothing but five stars carries almost no weight once their reviews are measured against their own baseline, which is what the normalization does. And the AI-adjusted percentile on every product page pulls thin samples toward the corpus mean, so a product measured by one or two reviewers sits near the middle there however glowing those reviews were. Separately, listings whose scraped figures fail our integrity checks are dropped rather than ranked — an implausible cluster of perfect star averages, or a rating count repeated verbatim across unrelated products, are collection artifacts rather than signal, and 16,368 of 70,958 rows were excluded on those grounds. Individual reviews are not audited: we read the aggregate Amazon publishes.
Are the buy links affiliate links?+
Yes. Each product links to its Amazon listing as an Amazon Associates link, and as an Amazon Associate we earn from qualifying purchases at no extra cost to you. The ranking is computed from the review data first and the link is attached afterwards — nothing about the score depends on where it points, or on whether you click it.
Do products have photos?+
A product shows its photo whenever we actually hold the image file, and a category-colored tile when we don't — the exporter only attaches a photo to a listing after checking the file is on disk, so a page never points at an image that isn't there. We're filling that library in now, and coverage climbs with each export. It will not reach every product — a portion of the stored image URLs no longer resolve — and it makes no difference to the ranking either way, which is computed from the review data alone.
Status: 1,509,012 products are live in the catalog across 14 categories, drawn from 1,509,012 Amazon listings we examined. 1,329,830 of the live rows currently carry a reviewer-normalized score and show a global and cohort percentile; the rest carry no rating at all and are marked as unscored until review coverage reaches them. The only listings excluded are ISBN-numbered ones, which belong to our Books catalog. See full methodology for the math.