Our Core Principles
These four principles guide every decision our Trust & Safety team makes.
Every Review is Real
We use a multi-layer verification system to ensure every review on FinanceReviews comes from a genuine person with real experience of the company they're reviewing.
Transparent Moderation
Our moderation guidelines are publicly documented. We apply them consistently and companies cannot pay to remove or alter negative reviews.
Independent & Impartial
FinanceReviews is not affiliated with any financial company. We do not accept payment to influence ratings, rankings, or review visibility.
Data Protection
We comply fully with GDPR and UK data protection law. Reviewer personal data is encrypted, never sold, and retained only as long as necessary.
Review Trust Levels
Not all reviews are equal. Our three-tier verification system ensures you always know how much weight to give each review.
Reviews submitted with name and email only. These undergo the most rigorous automated fraud screening and are held for moderation before publication.
Requirements
Reviews from users with a verified FinanceReviews account. Faster publication and higher weight in trust score calculations.
Requirements
Reviews from users who have completed full identity verification. These carry the highest weight in trust score calculations and are published immediately.
Requirements
AI Fraud Detection
Every review submitted to FinanceReviews is analysed by our 8-signal fraud detection system before publication.
Velocity Check
Detects unusually high review submission rates from a single user, IP address, or device fingerprint within a short time window.
Account Age Analysis
New accounts submitting reviews immediately after creation are flagged for additional scrutiny.
IP Clustering
Multiple reviews from the same IP address or subnet are identified and investigated for coordinated manipulation.
Content Similarity
TF-IDF cosine similarity analysis detects duplicate or near-duplicate review content across different accounts.
Sentiment vs Rating Mismatch
Our NLP model flags reviews where the written sentiment contradicts the star rating — a common sign of manipulation.
Pump & Dump Detection
Sudden spikes in positive reviews for a company are cross-referenced with market events to detect coordinated promotion.
KYC Identity Verification
Our highest trust tier requires government ID verification, making it significantly harder to submit fraudulent reviews at scale.
Behavioural Analysis
Mouse movement, typing patterns, and session behaviour are analysed to distinguish human reviewers from bots.
How We Handle Reports
When a review is flagged, here's exactly what happens.
Flag the Review
Any user can flag a review they believe violates our guidelines using the flag button on the review.
Automated Screening
Our AI system immediately re-analyses the flagged review against all fraud signals and guideline criteria.
Human Review
A member of our Trust & Safety team reviews the flag and the original review within 48 hours.
Decision & Action
We approve, edit, or remove the review based on our guidelines. Both the flagger and reviewer are notified of the outcome.
Appeal Process
Reviewers can appeal removal decisions. Appeals are reviewed by a senior Trust & Safety team member within 5 business days.
How Companies Can Engage With Their Reviews
Financial companies listed on FinanceReviews can claim their profile and respond to reviews. Here's what they can and cannot do:
Respond to reviews publicly
Companies can post a public response to any review, visible to all users.
Flag reviews that violate guidelines
Companies can flag reviews they believe violate our guidelines for human review.
Cannot pay to remove negative reviews
Reviews are only removed if they violate our guidelines — never for commercial reasons.
Cannot alter ratings or trust scores
Trust scores are calculated algorithmically and cannot be influenced by commercial relationships.
Claim Your Company Profile
Verified company representatives can claim their profile to respond to reviews, update company information, and access review analytics.