How research and verification work

Core principles

• Evidence before claims: Material factual statements need appropriate supporting sources.
• Transparency: Verified facts, platform-reported claims, independent findings, and editorial interpretation are kept clearly separate.
• Proportionality: The strength of a conclusion is matched to the quality and amount of available evidence.
• Recency: Information has to be current enough for the claim it supports.
• Corroboration: Important claims seek independent confirmation wherever that is practical.
• Fairness: Allegations are not treated as established facts, and missing evidence is not treated as proof of misconduct.
• Risk awareness: Material limitations and uncertainty are explained without exaggeration.
• Traceability: Every published review lists the sources we reviewed for it, with the dates we recorded for each one, and key facts on the page carry attribution labels saying whether a figure is the platform's own statement or comes from independent sources. A source that only repeats or summarises another source is shown as a restatement and is never counted as independent corroboration.
• No guarantees: Earnings, successful withdrawals, legitimacy, and safety are never promised.

What research covers

Where relevant to a platform, research may cover its official identity and operating website, the reward activity it offers, how users earn or qualify for rewards, reward currencies and payout methods, minimum payout thresholds and fees, geographic, age, and account eligibility, referral terms, withdrawal requirements, documented complaints or material concerns, and changes to how the platform operates. Not every field applies to every platform. Only relevant information supported by available evidence gets published.

Source quality and reliability

Sources are judged by relevance, authority, independence, specificity, and recency. No source category is automatically enough for every claim:

Claim-level verification

Material claims are assessed one by one instead of stamping an entire website with a single verification status. For each claim we record the statement, the sources supporting or challenging it, source type and authority, relevant dates, important limitations, and the resulting status with its rationale. What the published review shows is the outcome of that work: the reviewed-source list for the platform with the dates we recorded, and attribution labels on key facts saying where a figure comes from. The claim-by-claim working stays in our records rather than being displayed on the page, so a source list is not presented as proof of each individual claim.

A platform's own statement is attributed to the platform: it may support the wording “the platform states that…” without proving that the described outcome reliably occurs in practice. Unless a claim is independently corroborated, we present it as the platform's claim rather than as a finding.

A user report is attributed as a report, not as corroboration. It is labelled as reported by users until independent reporting or a reputation report supports the same claim, and a source that only repeats another source never counts twice.

When sources conflict

Disagreements between sources are weighed on four factors:

• Source quality: authority, expertise, direct access, transparency, and reliability.
• Recency: whether the information reflects how the platform operates today, including its current terms.
• Corroboration: whether independent evidence supports or challenges the claim. Multiple sites repeating one original source do not count as independent confirmation.
• Severity: what is at stake if the claim is wrong, especially for payout, financial, privacy, or account-access issues.

Official claims are not automatically preferred over credible independent evidence. Circumstances such as timeframe, jurisdiction, or payout method are checked before treating two sources as genuinely contradictory. Significant unresolved disagreements are described openly and labeled Conflicting evidence, and conclusions are reassessed when meaningful new evidence appears.

Payout information

Payout claims deserve particular care because people make financial decisions around them:

• A platform's advertised reward is kept separate from what has been independently demonstrated.
• Thresholds, fees, currencies, methods, and conditions are stated only when supported, and variation by country, account status, or payment method is checked.
• Screenshots and individual withdrawal reports are limited evidence unless adequately corroborated.
• One past withdrawal never implies future withdrawals, and promotional maximums are never presented as expected or representative earnings.
• Typical earnings are never estimated without defensible data and a clear method; unsupported payout details are omitted or labeled unconfirmed.
• Withdrawal limits and requirements, which cover the largest amount that can be withdrawn in a period, how often a withdrawal can be requested, and any conditions attached, are recorded next to the payout threshold, because they decide what a balance is actually worth.
• When the sources we reviewed do not state something, the review says so in plain words. Missing information is never shown as a zero, as no limit, or as an estimate, and figures that come from the platform are labelled as the platform's own.

Verification means documented against sources. We do not claim to have personally created accounts on, used, or withdrawn from any platform, and no such testing should ever be inferred from a label on this site.

Risk assessment

• Specific behaviors and issues are described rather than applying unsupported labels such as “scam”.
• Reports are attributed to their sources with unverified status marked, and serious allegations are corroborated where practical.
• An isolated report is distinguished from a documented pattern, and the absence of complaints is never presented as proof of reliability.
• Concerns are described in plain language rather than graded low, moderate, or high. We publish a severity rating only once a severity scale exists that we can apply consistently, and until then an ungraded concern is shown as ungraded.
• Material risks, limitations, and unresolved questions are stated plainly in every published review.

Evidence status labels

A published listing carries one of the assessed labels below. Research Pending is a separate state and is not an assessment. Color only reinforces the label; the text always says what it means:

No listing in the directory currently carries the Verified label.

Publication decisions

Publication is tiered rather than all-or-nothing. Each tier has a matching label, and the label is what a reader sees:

When a record is not published as a completed review

A record is not published as a completed review when any of the following is true:

• Material claims lack adequate supporting evidence.
• The available sources are materially unreliable for the claims being made.
• A serious unresolved issue could make the page misleading.
• Required source records or review dates are missing.

A platform is also taken down when it cannot be identified reliably, or when essential claims cannot be represented without misleading the reader. The reason is recorded internally and reconsidered if stronger evidence appears. Missing evidence is never proof of safety or wrongdoing in either direction.

Insufficient Evidence is a valid, clearly qualified publication status, not a euphemism for a refusal. It is the right answer when the platform is real and relevant but the evidence does not reach the threshold for a firmer conclusion, and it says nothing about whether the platform is fraudulent or legitimate.

Minimum requirements for publication

Alongside the editorial judgement above, each record is checked mechanically against a fixed list before its review page is produced. A record that fails any of these is treated as unpublished: it renders the same not-found page as an unknown URL, it is left out of the sitemap, and it is never offered in the directory's evidence filter.

• The record's editorial status is published, and it carries an assessed evidence status. A record still marked Research Pending is never published as a review.
• The record's own fields are valid: it has a unique identifier and slug, a website URL and any official-terms URL that are real http(s) addresses, at least one category that exists in the category registry, and readable creation, update, and check dates.
• The record has a genuine last-checked date: present, readable, and not in the future.
• The record has a valid page-updated date.
• At least one source record is attached to the record, and every source reference on it resolves to a source record with a usable http(s) URL.
• At least one claim is recorded for the record, every claim has at least one source behind it, and every source a claim uses is listed on the record.
• For a Verified record: every published claim is supported and dated, at least one claim is corroborated by independent reporting or a reputation report outside the platform (never a user report on its own, and never a source that merely repeats another source), and no recorded disagreement contradicts the status.
• A record with a material disagreement between credible sources is not labeled Insufficient Evidence, because the disagreement takes precedence.

What this site does not claim

• No blanket verdict on safety. We do not rate a platform safe or unsafe, and no evidence label should be read that way.
• No blanket verdict on legitimacy. A platform's own claims about who operates it, where it is registered, or that it is regulated appear as that platform's claims unless independently corroborated.
• No earnings promises, and no typical-earnings figures. Promotional maximums are never presented as what a user can expect.
• No scores or rankings. A Trust Score remains a Coming Soon placeholder; no numerical rating or simulated trust metric is published, and none will be until a scoring methodology is written down and published here.
• No traceability we cannot back. Every published review lists the source records behind it with the dates we recorded for them, and where a source only repeats another source that is stated plainly. A repeat is never counted as independent corroboration.
• No effect from money. Referral commissions, affiliate links, advertising, and sponsorship never decide whether a platform is listed, what evidence status it carries, what a review concludes, or whether a page is removed. Any such relationship is disclosed beside the link and on the listing.

Freshness and corrections

Every published review shows a last-checked date based on actual research, never bumped to make older content look current. Research is repeated when payout terms may have changed or when meaningful new evidence appears. A recent check date tells you when sources were reviewed; it does not independently verify every claim on the page. Conclusions are corrected or qualified whenever later evidence changes the assessment.

Two dates are kept apart on every review. The last-checked date is when the research behind the published claims was performed. The page-updated date records a change to the content of the page only. A fresh page-updated date never means the claims were rechecked, and the two dates are never merged into one.

A material correction is published on the review as a dated correction note, so a change of conclusion is visible rather than silent. Corrections are dated when they are published: we never backdate a correction to the day the original research happened, and we never invent a correction history for a record that has not had one.