RankRackObjective
Verification takes reasonable and proportionate steps to prevent and remove fake and incentivised reviews while protecting genuine critical feedback. "Verified" means the stated check was satisfied; it does not guarantee every factual claim, overall quality, or future provider performance.
Layered controls
Account controls: a confirmed email address before publishing, a date-of-birth check at sign-up (accounts are 18 or over), and one published review per account per provider.
Submission controls: the page must be open, the reviewer's account must not be on the provider's team, a stated service type, and a final confirmation step. Publication is immediate; the checks below run after the fact.
Risk signals: a new account posting an extreme rating, bursts of reviews from one account or on one page, invitation mismatches, near-duplicate text, reposts after a deletion, and reader reports. Signals are explainable rules; there is no device or network correlation and no third-party scoring.
Evidence check: a reviewer may supply order details - order and invoice numbers, purchase date, billing email - at submission. Staff check them privately; when they match, the review is marked Manually verified. We may also ask a reviewer for further confirmation through support.
Provider check: a provider may reply publicly, report a review or appeal it; it never sees the reviewer's evidence and cannot veto publication.
Human decision: a person considers the available information, alternative explanations, privacy and conflicts before anything moves.
Post-publication monitoring: new signals, reports and outcomes can trigger reassessment at any time.
Status labels
| Label | Meaning and weight |
|---|---|
| Verified purchase | RankRack confirmed the reviewer bought the service, through solid evidence or a supported integration. Full weight (1.0). |
| Manually verified | Staff matched the typed order details to the purchase. Weight 0.75. |
| Unverified | No qualifying check completed; this does not mean false. Weight 0.5. |
| Invited | A provenance mark, shown alongside the verification label, for a review submitted through a recorded provider invitation. It changes no weight, and the provider cannot edit or remove the review. |
| Under appeal | Shown while a provider's appeal is open. The review stays published while a person decides. |
The verification tier is the only thing that changes a review's weight. Incentivised reviews carry no label: the Guidelines for Reviewers prohibit them, and they are removed when found.
Independence and privacy
Providers cannot purchase verification outcomes. Staff declare conflicts, and a recused staff account cannot act on that provider; the platform's owners are not bound by recusal and their connections are published on the transparency page. No account of any kind can change a rating. Evidence access is restricted, never published, and retained according to the Privacy Policy's retention schedule - the order details themselves are removed 24 months after the check is decided. We collect only the typed order details listed above, never documents, government ID or payment-card data.
Fake-review prevention programme
We publish clear prohibitions on submitting, commissioning, procuring or misleadingly publishing fake or incentivised reviews, in the Guidelines for Reviewers and the Provider Guidelines.
We assess the risks in our platform, business model and user journey against known abuse patterns, and adjust the checks above as they change.
We use proportionate prevention, monitoring, reporting, investigation and sanctions, and review how well they work.
We present ratings and derived information accurately and explain the weighting rules where scores appear.
We do not suppress genuine negative reviews or let commercial relationships distort publication.