Ranking Criteria

How We Pick Top Rated, Best Value, Retailer Steal, and Price Low

Our picks are fully automated and data-driven. No sponsored placements, no manual curation. Every threshold is computed from the products actually in the category, so the bar scales correctly whether you are browsing $30 earbuds or $3,000 TVs.

Top Rated
  • Minimum 1,000 reviews: Products need enough real-world feedback to be trustworthy. Small review counts can be inflated or unreliable.
  • Above the median price floor: We compute the median price of all products in the category, then only consider products priced at 60% of that median or above. On a page with a $160 median, that means $96+. This stops a $39 no-name product from outranking a $348 Sony just because it has slightly higher stars.
  • Score: Rating x log(Reviews): A product with 4.5 stars and 60,000 reviews outranks one with 4.8 stars and 300 reviews. The logarithm prevents a product with 1 million reviews from having an outsized advantage.
  • Progressive fallback: If nothing qualifies under the strict rules (e.g. a new or niche category), the review threshold loosens to 200 so the section never goes empty.

Example: Headphones (median $160): minimum price floor is $96. Sony XM4 ($348, 62k reviews, 4.5 stars) scores higher than Soundcore Q20i ($39.99, 54k reviews, 4.6 stars) because the Q20i is below the price floor.

Best Value
  • At least 10% off list price: The discount must be meaningful as a percentage, not just a few cents off.
  • Minimum absolute dollar saving: We require savings of at least 8% of the category median price (minimum $5). On a $700 median TV page that is $56. This prevents cheap products with fake inflated list prices from gaming the metric.
  • Minimum 500 reviews: The product must be well-validated by real buyers before we surface it as a value pick.
  • Score: Savings% x log(Reviews): A 40% off product with 100 reviews loses to a 20% off product with 20,000 reviews. Both the deal quality and buyer validation are required.

Example: TVs (median $700): Best Value requires at least $56 off the list price, 10%+ discount, and 500+ reviews. A cheap product inflating its list price from $99 to $179 only to mark it 'down' to $99 will not qualify.

Retailer Steal
  • At least 2 retailers with valid prices: We need a market consensus to compare against. A product sold by only one retailer cannot have a meaningful price gap.
  • 25%+ price gap from the average of all other retailers: If five retailers sell a headphone at $400 and one sells it at $200, the gap is 50%. We flag the product when one retailer is 25% or more below the average of the rest.
  • Minimum absolute dollar gap: The gap must be at least 10% of the category median price (minimum $15). This prevents tiny-margin products from triggering the badge on a $3 difference.
  • Score: Gap% x log(Reviews): A no-name product with a 40% gap and 50 reviews loses to a well-reviewed product with a 30% gap and 5,000 reviews. User validation matters alongside the price gap.

Example: A Sony headphone is $400 at Amazon, Best Buy, Walmart and EcoFlow. Walmart then drops it to $220. The 45% gap qualifies it as a Retailer Steal, flagging it for shoppers who want the market outlier deal.

Price Low
  • At or below 2-year minimum price: We check 2 years of price history. The current price must be at or below the lowest price ever recorded (with 0.1% rounding tolerance).
  • At least 30 historical data points: Price history must be meaningful. A product with 3 data points cannot genuinely be called a price low.
  • At least 5% below 2-year average: The price must be a real standout. If a product has always been $199.99, it does not get the badge just for being at $199.

Example: Sony WH-1000XM4 gets a Price Low badge when its current price matches or beats its lowest ever tracked price and that price is 5%+ below its 2-year average.

No paid placements

Retailers and brands cannot pay to appear in our Top Picks. The algorithm runs entirely on public review counts, star ratings, list prices, and our own 2-year price history database.