Back to guides

How scores and rankings work

Every benchmark ranks its products with the same transparent method: each value gets a score from 0 to 100, the scores are averaged with the criteria weights, and the products are ranked by that average.

1. A score for each value

  • Rating (0 to 5 stars, half stars allowed): each star is worth 20 points, so 4 stars give 80.
  • Yes or no: yes gives 100, no gives 0.
  • Number: placed on a 0 to 100 scale between the minimum and maximum of the criterion (set by the author, or otherwise the lowest and highest values entered), in the direction set by the author (higher is better, or lower is better, as for a price). When every product has the same value, each gets 50.
  • List of options: the first option gives 100, the last gives 0, and the options in between are spread evenly.
  • Text, and numbers or lists marked as neutral, are shown for information and are not scored.

2. An overall score per product

  • The overall score is the weighted average of the product's scores on the scored criteria, from 0 to 100.
  • An empty value counts as 0: a product with missing data cannot rank higher just because it was compared on fewer criteria.
  • Products are ranked by overall score. Products with the same score share the same rank.

3. Weights you can change

Each criterion has a weight, from excluded to maximum importance. The author sets the default weights. When you view a benchmark, you can change them or exclude a criterion: the scores and the ranking update for you only, without changing the benchmark.

Where the values come from

Values are entered by the author of the benchmark, sometimes with AI assistance; the sources used are then listed on the product. The date of the last update is shown on every benchmark.