ReviewProof ⚖

What is the difference between a verified and an UNLINKED review?

By the Ingenious AI team · Published 14 November 2026 · Reviewed 14 November 2026

Run the free review audit See pricing · See a sample provenance report

Short answer

A verified review carries a link to a transaction record — an order, invoice or booking reference imported or attested with it — and its certificate shows that link. An UNLINKED review has no such link in the record. UNLINKED is not an accusation: it means the business cannot presently prove the transaction. ReviewProof keeps UNLINKED reviews in the audit and the risk report, and excludes them from the verified-experience widget, which displays only linked reviews in a kept or published state.

What is an UNLINKED review?

UNLINKED is ReviewProof's provenance label for a review with no verified-transaction link on record. The label describes the state of the evidence, not the character of the review: a heartfelt genuine review whose booking was taken by phone and never matched is UNLINKED, and so is a bought review with no transaction behind it at all. The label's honesty is the point — it separates 'we can show the transaction' from 'we cannot', without pretending to know more.

Why most healthy profiles have UNLINKED reviews

Real businesses collect reviews through messy channels: a platform prompt the customer answered without the business ever seeing an order number, a cash sale, a gift purchase reviewed by the recipient. A profile claiming a transaction link for every single review should prompt more questions than one with a candid linked/unlinked split. What matters is the trend and the effort: do you match links where records exist, import the reference when a review arrives through your own channels, and let the unmatched remainder stand honestly labelled instead of being quietly dropped from the story you tell?

ReviewProof computes every signal from the rows you import, never from a reputation guess about your business. The risk score is 100 × (gating × 1.0 + incentive × 1.0 + duplicate × 0.8 + AI-signal × 0.6) ÷ (2.6 × number of reviews), plus 15 points when a burst is detected, capped at 100. Each flagged review in the report names the imported row that triggered the signal, so a human can open that row, read it in context, and decide. The product labels these patterns signals for human review, not verdicts — and that restraint is deliberate, because a heuristic that overclaims would be its own kind of false evidence.

Raising the linked share without faking a single link

The legitimate levers are all in the plumbing. Export reviews from sources that carry order data. When you request reviews through your own email or receipt flow, include the transaction reference in the request record so the returning review can be matched. In the CSV you import, use the verified-transaction flag and reference columns rather than leaving them blank by habit. And for transaction-backed publishing, attach references properly: ReviewProof stores only a SHA-256 hash and a masked form of the reference, never the raw value, so publishing a provenance result does not expose customer data. What you must never do is invent references for old reviews — a fabricated link is worse than an honest UNLINKED label, and it poisons every true link beside it.

After the first audit, the habit that matters is cadence. Re-import on a regular rhythm — monthly suits most locations — so new reviews enter the provenance record while the underlying transactions are still easy to match. Each import logs provenance events, moderation starts from kept (no suppression), and the certificate for any review shows its chain: imported, transaction-linked or not, analysed with its signals, moderated with its decision. A record built gradually like this is worth far more in a dispute than one assembled in a panic the week the dispute arrives.

Mistakes that undo the work

Deleting UNLINKED reviews to raise the linked percentage is the cardinal error — it is suppression applied to your weakest-evidence rows, and the before-and-after record will show it. Backfilling plausible references is fraud, full stop. And describing UNLINKED reviews to customers as 'unverified, therefore fake' misstates your own system: some of your best genuine reviews will carry the label, and your staff should be able to explain why without flinching.

Two design choices keep the detection honest. First, nothing is scored until you import it: there is no scraping of your public profiles and no third-party score blended in. Second, the weights reflect what the FTC fake-review rule (16 CFR Part 465) actually prohibits — how a review was solicited weighs heaviest (gating and incentive markers at 1.0 each), while wording patterns weigh less (duplicates 0.8, AI-generation patterns 0.6). A burst of same-day reviews adds a flat 15 points because timing is visible in the dates you supply. You can reproduce every number in the report from your own file, which is exactly the point.

Worked example: a gift shop raises its linked share honestly

A fictional worked example, written to show the numbers and the decisions in practice.

A gift shop's first import of 156 reviews showed 48 linked and 108 UNLINKED, because most reviews arrived through a platform prompt the shop never saw orders for. The owner changed two things: online-order confirmations now carried the order reference into the shop's review-request record, and till staff stopped transcribing platform reviews into a spreadsheet (the step where references were being lost). Six monthly imports later the linked share on new reviews had risen from 31% to 78% — while the old UNLINKED rows remained in the audit, labelled as what they were. The widget's displayed rating barely moved; its provability transformed.

Your five-action checklist

  1. Read the verified and unverified counts in your report summary
  2. Trace where UNLINKED reviews enter — which channel loses the reference
  3. Add the transaction reference to your own review-request records
  4. Use the verified flag and reference columns on every import
  5. Never invent, borrow or backfill a reference for a review

Frequently asked questions

Does UNLINKED mean a review is fake?

No. It means no transaction link was imported for it. Genuine reviews are UNLINKED whenever the business cannot match the review to its records.

Can an UNLINKED review appear in the widget?

No. The verified-experience widget only displays reviews that are both transaction-linked and in a kept or published moderation state.

How is a transaction reference stored for published results?

Only a SHA-256 hash and a masked form are held; the raw reference is never stored, and email-shaped references are refused outright.

Can I link a review later, when I find the order?

Yes — attaching a reference later is a provenance event, logged in the chain like any other. Finding and matching real records is exactly the habit the label rewards.

Keep reading

Related guides from the ReviewProof library:

Start with the tool itself: open the free audit desk, see how ReviewProof works, compare Free, Pro and Agency plans, inspect the sample provenance report, or browse all ReviewProof guides.