What Is a Proof-Based Review?

A proof-based review is a rating or written assessment of a business that is backed by verifiable evidence of a real interaction — an order, a contract, an invoice, a completed escrow transaction, or another artifact that can be checked. Instead of asking you to take a stranger's word for it, a proof-based review ties the opinion to something that actually happened. On Trusgent (准策) this is a core trust primitive: reviews are not free-floating opinions, they are claims attached to proof.

This entry explains what counts as proof, how proof-based reviews differ from ordinary or anonymous reviews, why they resist manipulation, how proof is attached while keeping sensitive details private, how verified-transaction reviews feed the Trust Score, and the honest limits of the model.

The Problem With Ordinary Reviews

Most review systems on the open web share the same weakness: anyone can post, and almost nothing is checked. A five-star rave can be written by the owner's cousin. A one-star attack can come from a competitor who never bought anything. Review farms sell bulk positive ratings; extortion rings threaten negative ones. Because the underlying interaction is invisible, the reader cannot tell a genuine customer from a paid actor.

This matters more than ever now that AI agents read reviews on a buyer's behalf. An agent shortlisting vendors cannot afford to weigh fabricated text the same as real experience. Unstructured, unverifiable star ratings are exactly the kind of signal an agent should discount — yet historically there has been no machine-readable way to tell which reviews are real.

Proof-based reviews exist to close that gap. They follow the Trusgent principle of evidence over claims: a review is only as strong as the proof behind it.

Supported Proof Types

Trusgent supports several categories of evidence, each attached to a review to substantiate that a real interaction occurred:

•Order: a record of a purchase or booking — an order confirmation, order number, or platform receipt showing the reviewer transacted with the business.

•Contract: a signed agreement or statement of work establishing an engagement between the two parties.

•Invoice: a billing document showing services rendered or goods supplied, linking payment to the relationship.

•Screenshot: captured evidence of a conversation, delivery confirmation, dashboard, or other interaction that documents the engagement.

•Manual verification: a Trusgent reviewer or trusted process inspects supporting material and confirms the interaction when automated checks are not enough.

•Escrow-completed transaction: the strongest form — a deal that ran through Trusgent's escrow and settled, so the platform itself witnessed the transaction end to end.

These types form a ladder of confidence. A screenshot is helpful but weaker; an escrow-completed transaction is the gold standard because Trusgent observed the money move and the deal close, leaving little room for invention.

How Proof Is Attached, and What Stays Private

When a reviewer submits a proof-based review, they attach the relevant artifact. The verification layer checks that the proof is consistent — that the order, invoice, or contract genuinely connects the reviewer to the business being reviewed. Once confirmed, the review is marked with the proof type it carries, so any reader can see the basis of the rating.

Crucially, attaching proof does not mean publishing it. The principle is transparency about the existence and type of evidence, not exposure of its contents. So:

•The fact that a review is backed by, say, an escrow-completed transaction is public.

•The private contents — invoice amounts, contract clauses, personal contact details, internal screenshots — are not put on display.

•Sensitive figures and identities are kept confidential and used only to substantiate that the interaction was real.

This lets a reader trust the verification without forcing either party to leak commercial secrets. Human control is preserved: a business is not stripped of confidentiality just because a customer reviews it.

Why Proof-Based Reviews Resist Manipulation

The manipulation that plagues ordinary reviews depends on the interaction being invisible. Proof-based reviews remove that cover:

•Fake positives are harder. To post a glowing review you must show evidence of a real order, contract, or settled deal — not merely create an account.

•Fake negatives are harder. A competitor or extortionist who never transacted cannot produce proof, so their attack carries no verified weight.

•Volume attacks lose power. Buying a thousand reviews means forging a thousand checkable artifacts, which is far costlier than spinning up a thousand anonymous accounts.

•Ranking stays honest. Because the Trust Score weights verified evidence and cannot be bought, flooding the system with unverified praise does not move it.

The result is a signal that is expensive to fake and cheap to trust — the opposite of the open-web norm.

How Verified-Transaction Reviews Feed the Trust Score

Not all reviews count equally toward a business's Trust Score. Reviews backed by stronger proof — especially escrow-completed transactions — carry more weight than weaker or unverified input. This is deliberate. The Trust Score is meant to summarize confirmed standing, so it leans on the evidence the platform can stand behind.

In practice this means a steady stream of verified-transaction reviews builds Trust Score over time, while unverifiable noise has little effect. Because the score is transparent rather than a black box, a business and its counterparts can understand why standing rose or fell, and a buyer's agent can reason about it instead of guessing. And since the score cannot be bought, the only durable way to raise it is to do real, verifiable business well.

Honest Limits

Proof-based reviews improve transparency; they do not make every statement perfect:

•Proof confirms that an interaction happened, not that every opinion about it is fair or representative.

•A verified customer can still hold a subjective view others would disagree with.

•Some legitimate interactions are hard to evidence, so absence of proof is not proof of fraud.

•Verification reduces manipulation but cannot eliminate every edge case.

The honest claim is narrow and therefore credible: a proof-based review tells you the reviewer really dealt with this business, and how that was confirmed. It does not promise that the review is the final word.

How Buyers and AI Agents Should Read Them

•Look at the proof type first. An escrow-completed or invoice-backed review deserves more weight than an unverified one.

•Read the strength of evidence, not just the star count. A few verified reviews can outweigh many anonymous ones.

•For agents: prefer reviews with machine-readable proof markers, weight them by proof strength, and discount unverified text.

•Combine reviews with the broader Trust Passport — credentials, Trust Score, and badges — rather than reading any single signal in isolation.

Related Terms

•Trust Score — the transparent, unbuyable standing that verified reviews feed.

•Agent-ready Trust Passport — the verifiable identity that hosts your reviews.

•Trust Badge — an embeddable mark reflecting confirmed standing.

•Escrow-completed transaction — the strongest proof type, witnessed by the platform.

•Trusgent ID — the stable identifier that ties reviews to the right entity.

FAQ

•Can I review a business I never dealt with? No. Without proof of a real interaction, a review carries no verified weight.

•Is my invoice or contract made public? No. Only the existence and type of proof is shown; the contents stay private.

•Do negative reviews need proof too? Yes. The same evidence requirement applies, which is what blocks baseless attacks.

•Does more reviews always mean a higher Trust Score? No. Verified, stronger-proof reviews count more; unverified volume does little.

•Can I pay to improve my reviews or score? No. The Trust Score cannot be bought; only real, verifiable business builds it.