RatingHub
Trust & Integrity

Editorial Policy

How RatingHub ensures reviews are genuine, scores are honest, and businesses cannot buy their way to a better rating.

Last updated: 19 June 2026

RatingHub is funded by advertising and premium business listings — not by charging businesses to improve or suppress reviews. Our editorial independence from advertiser relationships is absolute and non-negotiable.

Who Can Write Reviews

Any registered user who has had a genuine experience with a business, product, or service may write a review. You do not need to prove purchase — but you must have personally experienced the entity you are reviewing.

Business owners, employees, and anyone with a direct financial relationship to a company may not review that company. Our AI checks for reviewer patterns consistent with insider activity.

Review content must be original and written by the submitting user. Copy-pasting from other platforms or AI-generated review bodies submitted verbatim are not permitted.

How We Detect Fake Reviews

Every review submitted to RatingHub is automatically scored by our AI manipulation detection system before it can be published. This analysis checks for: unusual language patterns, multiple submissions from similar IP addresses, rating distribution inconsistent with reviewer history, and timing patterns associated with coordinated campaigns.

Reviews with a manipulation score above 60/100 are routed to the moderation queue for human review before publication. Reviews above 80 are automatically held pending a moderator decision.

We publish each review's AI verification status openly on the review card. Nothing is hidden.

Verified Purchase Badges

A 'Verified' badge appears on reviews where the reviewer has uploaded a purchase receipt or where a connected business has confirmed the purchase via our webhook integration.

Verified reviews carry 1.5x weight in DNA score calculations. This gives weight to customers who demonstrably spent money, while still allowing unverified genuine experiences to count.

Verification is handled server-side. Uploaded receipts are analysed by Gemini Vision and not stored beyond the verification check.

Moderation Process

Our moderation queue is staffed by trained human moderators who review flagged content. Moderators may approve, reject, or escalate reviews. All moderation actions are logged in our audit trail.

Businesses may flag reviews they believe violate our guidelines, but flagging does not remove a review. Only a moderator decision can change a review's status.

Businesses cannot pay to remove, suppress, or improve reviews. Reported reviews that do not violate our guidelines are reinstated. This is non-negotiable.

DNA Scoring Methodology

The DNA Score is a composite rating across five axes — Performance, Value, Reliability, Support, and Experience — each scored 0–100 by reviewers. The composite is a weighted average: Performance 25%, Value 20%, Reliability 25%, Support 15%, Experience 15%.

Scores are calculated from all published reviews. Flagged, rejected, or under-review content does not count. The score updates in real time as reviews are approved.

We do not adjust, smooth, or editorially influence DNA scores. What the community rates is what appears.

AI-Assisted Features

RatingHub uses Google Gemini AI for three opt-in features: (1) review enhancement — users may request an AI-polished version of their written review, but the original version is preserved and used by default; (2) company description generation — used to fill stubs when a business hasn't claimed their listing; (3) AI Summary — a synthesised summary of recent reviews shown on company profiles, clearly labelled as AI-generated.

AI-generated content is always clearly labelled. Scores are never set by AI — only by human reviewers.

Our Gemini models are grounded with web search for entity discovery and do not fabricate company information. We verify sources before publishing enriched data.

Questions about our editorial policy? Contact us or email hello@theratinghub.com

See also: Review Guidelines · Privacy Policy · Terms of Service