We Rank it: by re-scaling reviews by each user (z-scores)
Over 1 Billion reviews utilized
Our rankings remove bots, decrease weight for only 5-stars rating users, remove single review users, and more — to get the REAL insights.
Product categories
Example of why our rating system is the best:
| Hotel | Dan | Ari | Mel | Jen | Tim | Avg |
|---|---|---|---|---|---|---|
| A | 5/5 | — | 5/5 | 3/5 | 3/5 | 4.0 |
| B | 5/5 | 5/5 | 5/5 | 3/5 | 3/5 | 4.2 |
| C | — | — | — | 4/5 | 4/5 | 4.0 |
| D | 4/5 | 5/5 | — | 1/5 | 2/5 | 3.0 |
| Least important | Most important information | |||||
− each reviewer’s own average
| Hotel | Dan | Ari | Mel | Jen | Tim | Avg |
|---|---|---|---|---|---|---|
| A | +0.33 | — | 0.00 | +0.25 | 0.00 | +0.15 z +0.02 |
| B | +0.33 | 0.00 | 0.00 | +0.25 | 0.00 | +0.12 z −0.02 |
| C | — | — | — | +1.25 | +1.00 | +1.13 z +1.42 |
| D | −0.67 | 0.00 | — | −1.75 | −1.00 | −0.85 z −1.41 |
Summary
raw score average
Follow Hotel B: ringed at 4.2, first on the raw average — and underlined at 49, third once every reviewer is re-centred.
Other sites would have you choose Hotel B. Rankquant clearly shows you math-perfected Hotel C is superior.
The process
Three deterministic steps from review to rank
Per-reviewer z-score
Subtract the reviewer's personal mean, divide by their personal stddev. Every reviewer now speaks the same dimensionless language.
Three lenses
Unweighted (headline) — every qualifying reviewer counts the same. In cohort, because a $500 red is not a $10 white. AI-adjusted, which discounts a score for how much confidence its sample actually supports, pulling thin samples toward the average.
Percentile
Raw scores are not on a 0–100 scale; the percentile puts them there. Cohort percentile re-ranks within the same category to give you a normalized scoring system.
Why us
What we do that Google and ChatGPT can't
Search ranks by links and click behavior. We rank by what bias-corrected reviewers actually thought. The most-clicked wine isn't the highest-quality wine.
AI summarizes whatever it scraped — usually flat averages or vibes. Our methodology is published, deterministic, and reproducible. Every score has a paper trail.
Native platforms have rating clusters: 4.5 means nothing when every product gets a 4.5. We strip the cluster effect by re-centering each reviewer on their own scale.