The Gauge Methodology: How Every Score Is Calculated

Why we score this way

We don’t claim to have worn forty fragrances through a heatwave or washed twelve t-shirts a hundred times — publications that make such claims at scale rarely have. What actually exists, in enormous volume, is the experience of people who bought the product: thousands of verified owners reporting how the collar held, whether the scent survived lunch, if the watch kept time. Our job is to aggregate that experience rigorously, weight it honestly, and publish the method so you can check our work.

The four components

Owner sentiment — 40%. Verified-purchase reviews aggregated across major retailers, weighted by recency and volume, with duplicate and incentivized reviews filtered. One brand’s own site never dominates.

Spec analysis — 30%. Products are normalized against category benchmarks: fabric weight and fiber staple for clothing, movement and case specs for watches, ingredient lists and concentrations for grooming. Marketing adjectives score zero.

Price integrity — 20%. Twelve months of price tracking. A product that lives permanently “on sale” from an inflated anchor price loses points; stable honest pricing gains them.

Durability signals — 10%. Public return-rate data, warranty terms, repairability (resoleable, serviceable, refillable) where disclosed.

The four things we never do

We never compute a score after placing an affiliate link — scores are locked first, links added second, and the highest scorer is the highest scorer whether or not it pays a commission. We never accept payment for a score, a ranking position, or a review. We never present aggregation as personal testing. And we never let a score sit stale: every score carries its calculation date and a quarterly re-score cycle, because products get reformulated, re-priced and quietly downgraded.

What the numbers mean

9.0+ — exceptional; buy-once territory. 8.0–8.9 — strong; the best in most categories live here. 7.0–7.9 — good with caveats we’ll name. Below 7 — we’ll tell you what to buy instead.

We publish low scores; a rating system that only says nice things isn’t a rating system.

The receipts

Where a guide leans on numbers — prices, resole costs, review counts, alteration rates — it carries a receipts panel: every figure with its named source, a citation link, and the date we checked it. Three rules govern the panel. Figures are list prices from the source page on the check date, not cherry-picked promos. Links in a receipts table are citations only — never affiliate links. And receipts are rechecked at every guide review, on the same dated cycle as the scores; when reality moves, the guide’s math is recomputed rather than defended. If a receipts row and reality disagree, use the correction channel above — a documented error triggers a re-check, logged publicly.