Methodology
How an Index score is calculated
Every Index score is the same simple sum. This page explains the parts, where the numbers come from and what we do when the evidence is thin.
Effective 1 October 2026
At a glance
- 4 to 6 criteria per category, with weights that add up to 100%
- Each product scored 0 to 100 on each criterion
- Index score = sum of (criterion score × weight), rounded
- Weights and sub-scores are printed on every report
1. Choosing the criteria
We start with what owners complain about and praise most often in the category. For headphones that is noise cancelling, sound, comfort, battery and calls. For a video doorbell it includes the yearly cost of the subscription, because a cheap doorbell with an expensive plan is not cheap.
2. Setting the weights
The weights reflect how much each criterion matters to a typical buyer. They are set before we score any product, so they cannot be tuned to favour one. If we change the weights in a report, we say so at the top of the report.
3. Scoring each criterion
Each sub-score from 0 to 100 draws on the evidence we can verify:
- Published specifications from the maker, such as battery life, weight or resolution.
- Independent measurements from named labs and publications, cited in the report’s references.
- Owner reviews pooled from large retailers, where we look for faults that recur across hundreds of reviews.
- Running costs such as subscriptions, replacement heads and accessories, from the maker’s own price list.
- Price at the date shown, for the value criterion.
Scores are relative to the products available in that market today. A 90 means near the best you can currently buy on that measure, not perfection.
4. Adding it up
| Criterion | Score | Weight | Points |
|---|---|---|---|
| Example A | 90 | 30% | 27.0 |
| Example B | 80 | 25% | 20.0 |
| Example C | 70 | 25% | 17.5 |
| Example D | 85 | 20% | 17.0 |
| Index score | 100% | 82 |
5. Keeping scores current
We re-check prices monthly and review each report in full at least twice a year. When a new model launches, older products are re-scored against it, so a score can go down without the product changing.
What we do not do
- We do not round scores up to award a badge.
- We do not score products we could not find reliable evidence for. These are listed as “considered, not scored in full”.
- We do not let price alone decide first place.