Home › Methodology
How We Score: The JewelryScore Methodology
Every brand on Jewel Meter receives a JewelryScore — a letter grade from A+ to F derived from a 100-point scale. Here's exactly how we calculate it.
The Three Pillars
Every JewelryScore is composed of three weighted pillars. The weights reflect what matters most to jewelry buyers based on our analysis of purchase decision factors across thousands of customer reviews.
Quality — 40 Points
The largest pillar measures what customers say about craftsmanship, durability, tarnish resistance, comfort, and how the piece holds up over months of daily wear. We aggregate sentiment from Reddit discussions, Google Reviews, and Trustpilot reviews, weighting long-term ownership feedback more heavily than initial impressions.
Data sources: Reddit, Google Reviews, Trustpilot
Value — 35 Points
Value is scored within the brand's price tier (under $50, $50–150, $150–500, $500+). A $30 name necklace is not penalized for using different materials than a $300 one — it's compared against other options in the same range. We evaluate price relative to materials, comparable pieces from competitors, and customer satisfaction-per-dollar.
Data sources: Product pricing, materials specifications, competitive comparison
Style — 25 Points
The style pillar measures trend alignment and aesthetic range. We track social media mention velocity on TikTok and Instagram, Pinterest save rates, and coverage from style publications. Brands offering broader aesthetic range within their categories score higher.
Data sources: TikTok/Instagram trend volume, Pinterest saves, style publication mentions
Letter Grades
| Grade | Score Range | What It Means |
|---|---|---|
| A+ | 95–100 | Exceptional across all pillars |
| A | 90–94 | Excellent overall with minor trade-offs |
| A− | 85–89 | Very strong with some category-specific weaknesses |
| B+ | 80–84 | Good option with notable strengths |
| B | 73–79 | Solid but with meaningful limitations |
| B− | 67–72 | Acceptable with significant trade-offs |
| C+ | 60–66 | Below average in one or more pillars |
| C or below | <60 | Not recommended based on current data |
Data Collection Process
Every month, for each brand we cover, we run the following process:
- Scrape: Automated pull of top 100 Reddit comments, Google Reviews, and Trustpilot reviews mentioning the brand. Instagram and TikTok mentions are captured as supplementary data for the Style pillar.
- Summarize: Raw data is processed through our AI analysis pipeline to generate a Social Verdict — identifying key praise themes, recurring complaints, overall sentiment score, and representative quotes.
- Score: Pillar scores are calculated from the processed data using our weighted rubric.
- Review: A human editor reviews every score and summary for accuracy, tone, and fairness.
- Publish: Updated scores and Social Verdicts are published with the update date clearly displayed.
Integrity Rules
- Scores are produced by the rubric, never predetermined.
- All brands receive honest assessments, including real trade-offs for top scorers (e.g., longer shipping for customized pieces, no physical retail, limited fine jewelry selection).
- Coverage depth is naturally asymmetric — top-scoring brands earn more detailed treatment because there's more to say about what they do well.
- This methodology page is always public and always up to date.
Questions About Our Methodology?
We believe in full transparency. If you have questions about how we calculate scores or how we source our data, reach out to our editorial team.