How We Score Products

Our Methodology

Every product on this site receives a composite score from 0 to 100, calculated from five weighted dimensions. We believe that a single star rating doesn't tell the full story — so we built an algorithm that factors in what actually matters to buyers.

Scoring Dimensions

Rating (35%)

The product's average customer star rating on Amazon, scaled linearly from 1–5 stars to a 0–100 score. This is the single most important signal of product quality.

Review Volume (25%)

The number of customer reviews, log-scaled to prevent mega-products with tens of thousands of reviews from unfairly dominating. A product with 500 genuine reviews can score nearly as well as one with 5,000.

Value (20%)

How the product's price compares to others in its category. The most affordable products score highest. We rank by price within the category, so a $300 robot vacuum competes against other robot vacuums, not against $30 accessories.

Features (10%)

The richness of the product's feature set, measured by the number and detail of listed features. Products with comprehensive feature descriptions score higher, as this often correlates with manufacturer investment in the product.

Recency (10%)

How recently the product appeared in our data. Newer products get a slight boost, with a 30-day decay. This helps surface fresh products while still rewarding proven performers.

Data Freshness

Our pipeline runs daily, pulling the latest prices, ratings, and availability from Amazon. Scores are recalculated every day to reflect current data. Every review page shows when it was last updated.

AI-Assisted Content

The introductory buying guides and product summaries on this site are generated with the assistance of AI (Anthropic Claude). We use AI to synthesize product data into helpful, readable content — but the scoring algorithm and product rankings are purely data-driven.

Affiliate Disclosure