Rating 50% Volume 50%
A handful of excellent reviews counts — but cannot carry the score. With thin data, the star average leads.
Volume signal saturates at 100 reviews.
Every product on PickSmartly earns a PM Score from 0–10. Most sites keep that number in a black box — here is the exact formula, every weight, and every tier that produces it.
Two public Amazon signals go in; one number comes out. The customer rating measures quality; the review count measures consensus — how many people are standing behind that rating.
PM Score = wQ × Quality + wC × Consensus
The weights are not fixed: they shift with review volume, in three tiers. The fewer reviews exist, the more the star average has to say; once tens of thousands have weighed in, the crowd is the evidence.
Each product is scored in exactly one phase, decided by its review count.
Rating 50% Volume 50%
A handful of excellent reviews counts — but cannot carry the score. With thin data, the star average leads.
Volume signal saturates at 100 reviews.
Rating 30% Volume 70%
The crowd has spoken. Consensus now dominates, and a lone wave of early hype can no longer prop a product up.
Volume signal saturates at 1,000 reviews.
Rating 10% Volume 90%
At this scale the volume signal saturates at maximum. A 4.2★ across 25,000 reviews is what the world actually receives.
Volume signal saturates at 1,000 reviews.
One product, one rating — three phases. The same 4.5☆ earns a higher score as consensus grows, while the star average itself matters less and less. Every value below is computed live by the same formula the rankings use.
Why 8.5 → 9.7 → 9.9 for the same stars? Quality contributes 4.5, then 2.7, then 0.9 — as consensus firms up, the crowd’s verdict takes over the score.
Full arithmetic for JCKLW 49065-0724 49065-7010 Oil Filter Compatible with Kawasaki Club Car 1016467 41016467 Kubota T1460 T1560 T1700H T1770 T1870 TG1860G Robin EH18V EH64 EH65 Yamaha MX775 MX800 MX825 2pcs, recomputed live from current data.
0.0 × 0.5 + 2.2 × 0.5 = 1.10
A $19 kettle and a $190 kettle face the identical rubric. Price is a purchase decision, not a quality signal — and Amazon prices move too fast to score against.
Both inputs — star rating and review count — come from Amazon’s own public data. Nothing proprietary, nothing self-reported by brands.
The same equation scores every product on the site, with no per-brand overrides, no manual boosts, and no exceptions — including the ones on this page.