How We Test, Rank, and Render Verdicts
Every review on this publication applies the same five-step framework. The framework defines what evidence counts, how it is weighted, and how a verdict is reached. The per-review methodology in each article is this framework applied to one category, published alongside the ranking.
Step 1. Define the criteria
Before any product is evaluated, the criteria for the category are defined and written down. Criteria are observable and tied to what the product is supposed to do.
For a watch: movement type and caliber, case and crystal material, water resistance stated to ISO 22810, bracelet or strap quality, and warranty. For a leather good: tannage, grain class, hardware, stitch density, and edge finishing. For jewelry: stone identity and grade, metal and plating, build, and the terms of any guarantee. The categories differ. The discipline does not. Criteria are published with the review and applied uniformly to every entry.
Step 2. Collect the evidence
For each entry, the team collects the strongest available evidence against each criterion. The hierarchy:
- Third-party reports against named standards, such as a gemological report on a recognized grading scale or a movement specification from the named caliber maker.
- Accredited material and performance data, such as a water-resistance rating tested to ISO 22810 or a leather classification consistent with the trade definitions.
- Supplier and component disclosures from named movement makers, tanneries, or stone sources that confirm the relationship.
- Manufacturer specifications, treated as supporting evidence when corroborated by the sources above and treated with caution when standing alone.
- First-hand inspection of build, finish, weight, and hardware where the team can handle the product.
A claim that rests on evidence the team cannot verify is recorded as unsubstantiated, and the review identifies it as such.
Step 3. Score against the criteria
Each entry is scored against each published criterion using the evidence collected. The scoring approach is documented in the per-review methodology and varies by category. For some categories the criteria are weighted equally. For others, where one criterion is load-bearing, such as water resistance on a dive watch, the weights are stated explicitly. Scoring happens before the ranking is composed. The score feeds the ranking; the score is not adjusted after the fact to produce a preferred order.
Step 4. Rank from the scores
The ranking is the score order, from the strongest evidence-substantiated performance to the weakest. Ties are reported as ties. The team does not invent a differentiating criterion after the fact to break one. Where the gap between adjacent entries is small relative to the uncertainty in the evidence, the review says the entries are functionally equivalent on the criteria as applied.
Step 5. Render the verdict
The verdict for each entry is a short editorial paragraph: what the entry does well, where the evidence is thin, and which reader it suits. Verdicts use plain language and avoid the marketing register the brands themselves use. Where an entry cannot be fully substantiated but has real strengths, the verdict names both, so a reader can weigh them.
What the methodology does not do
The methodology does not certify products. It is not a regulatory standard. It does not make health, safety, or financial claims about any specific product, and it does not substitute for professional advice. It produces a public, reproducible editorial ranking based on the published evidence. That is the scope.
Per-review methodology
Every published review carries its category-specific methodology in a dedicated section, so a reader can see the criteria, the evidence rules, and the scoring approach used for that review rather than only the abstract framework above. Published reviews live at /reviews.