Trust you can actually trust.
A review and trust system for African businesses, NGOs, schools, and clinics, engineered against fakes, self-praise, and brigading, and built to protect the people who leave reviews.
A TrustScore weighted for recency and verification, not a vanity average
Self-reviews blocked at the database; a fraud engine watches for brigading
Reviewers stay private: the raw reviews table is never publicly readable
Reputation, done honestly
Reviews only mean something if they are real. Every part of Ardent Reviews is built to keep them that way.
TrustScore, not a vanity number
A score from 0 to 100 weighted for recency and verification, shown with an honest 'still gathering data' band. It is never one opaque average that a handful of reviews can swing.
AI 'what people say'
A plain-language summary of the themes across every review, so anyone can grasp the reputation of a place at a glance: the good and the bad.
Claim & respond
Owners verify their listing by KYC, edit their public profile (logo, contact, socials), and reply to any review with a right of reply. Honest feedback stays, but you get a voice.
Verified invitations
Invite real customers to review in batches so your rating reflects everyone's experience, not just the loudest voices or the angriest day.
Anti-gaming by design
Self-reviews are blocked at the database. A fraud engine flags IP velocity, duplicate text, new-account bursts, and coordinated rating swings against admin-tunable rules.
Private & portable
The base reviews table is never anon-readable. A public API, an embeddable rating widget, and JSON-LD let a business show its verified trust anywhere on the web.
How trust is built
01
Find or add an entity
Search businesses, NGOs, schools, and clinics, or add the one you want to review.
02
Write a star review
Rate overall and by dimension, and mark a verified experience where you genuinely have one.
03
Owners claim & reply
The business verifies ownership by KYC, edits its profile, and responds to feedback.
04
Trust compounds
The TrustScore updates continuously, weighted for recency and verification, with a confidence band.
Verified, PII-safe, and gaming-resistant, including for us.
Ardent Reviews is designed so that a rating is hard to fake and honest to read. It even holds Ardent itself to the same standard.
Self-reviews are blocked at the database level; a fraud engine watches IP velocity, duplicate text, and coordinated rating swings.
The raw reviews table is never anon-readable: every public read goes through a PII-safe projection.
Guests can leave a verified-receipt review after a real donation, signature, or ticket, identified by a salted hash, not an email.
Even Ardent reviews itself: two pinned self-entities held to maximum moderation, with every review human-approved.
Owners get a right of reply and can report policy-breaking reviews, but can never delete honest criticism.
The TrustScore shows a 'still gathering data' band when evidence is thin: it never fakes confidence it hasn't earned.
Questions, answered
No to all three. Self-reviews are hard-blocked at the database level, so a business literally cannot rate its own listing. Ratings are never for sale. And owners can reply to any review or report ones that break the rules for moderation, but they can never delete honest criticism. The whole point is that the record is not the owner's to edit.
A star average is trivially gamed: a burst of five-star reviews on a quiet listing swings it overnight. The TrustScore is a 0-to-100 measure weighted for recency and verification, shown with a confidence band so a thin record reads as 'still gathering data' rather than a fake, precise number. It rewards a steady history of verified experiences over any one-off spike.
Layered defence. A fraud engine scores IP velocity, duplicate text, new-account bursts, and coordinated entity-level rating swings against rules that admins can tune. AI pre-publish moderation checks content before anything goes live. And verified-experience reviews are weighted more heavily, so an organised pile-on of anonymous ratings moves the score far less than genuine, receipted feedback.
Signed-in users can review any entity. Guests can leave a verified-receipt review right after donating, signing a petition, or attending an event, identified by a privacy-safe salted hash rather than a stored email. Either way the raw reviews table is never publicly readable, so your identity is not exposed by leaving honest feedback.
Yes. Reviews are built to be portable: a public API, an embeddable rating widget, and JSON-LD structured data let you display your verified TrustScore and reviews on your own pages and in search results, without anyone being able to fake the number behind it.
Businesses, NGOs, schools, and healthcare facilities today. Government and religious entity types are fully modelled in the system but ship disabled for now, pending a legal and brigading-risk assessment, because those categories carry a higher risk of coordinated abuse.
Show the world you can be trusted.
Read verified reviews, claim your listing, and put a TrustScore you earned honestly on your own pages. Trust that is real, and provable.