How it works

Proof labelled. Opinion independent.

We label the evidence behind each reviewer, detect coordinated manipulation, and keep companies outside the publication decision.

1Email ownershipVerified address
2Proof levelIdentity, invite or transaction
3Integrity checksBehavior and duplication
4Automatic outcomePublish, hold or reject
01 · Reviewer

Share a first-hand experience

Verify your email and describe what happened. An order or case reference is optional and always private.

02 · Proof level

Show what was verified

Identity verification is the minimum. A one-time company invitation or optional signed transaction match earns a stronger public badge and a different risk threshold.

03 · Integrity

Check behavior, not sentiment

The engine scores velocity, device and network patterns, repeated content, account reputation, suspicious links, private data, conflicts and blocked indicators. Positive and negative ratings use identical rules.

04 · Company

Reply—without controlling publication

A company self-verifies with work email plus DNS TXT control. It may optionally import transactions, invite customers, reply and report a precise concern, but cannot change ratings or approve reviews.

05 · Automation

Deterministic, auditable outcomes

Lower proof receives longer holds and stricter publish limits. Critical signals or blocking clusters reject, while every result records its policy version.

06 · Recheck

Correct changed evidence automatically

A reviewer can request a rate-limited recheck. Late transaction imports can upgrade the proof level, while cluster changes schedule linked reviews for another pass.

Honest limitation: identity, invitation and transaction evidence prove different things. No fully automated platform can guarantee subjective narrative accuracy. The visible labels prevent stronger claims than the evidence supports.