How is the overall reliability score calculated?

It is the mean of four equally-weighted dimensions — paid on time, brief clarity, responsiveness, and would-work-again — each rated 1 to 5 by creators after a paid deal, averaged across all of a brand’s published reports.

Why is a minimum of three reports required?

A single report can’t be a fair public verdict and could identify the creator. Requiring three reports protects anonymity and converts anecdote into a reliable signal.

How are fake or malicious reports handled?

Reports pass a moderation gate before publishing, must reflect a real deal, and stick to verifiable facts. Brands get a right of reply, and reports that violate the guidelines are removed.

Methodology

How the reliability score works

Transparency is the product. Here is exactly how a brand’s score is built, protected, and kept fair.

The four dimensions

Every report rates four things, each on a 1–5 scale. They are weighted equally — reliability is not one number with caveats, it is four honest answers.

Paid on time. Did the brand pay the agreed amount, on the agreed date? (Invoices clear on schedule, no chasing.)
Clear brief. Was the scope clear and stable, or did it creep? (Deliverables were defined up front and held steady.)
Responsive. Did the brand reply in reasonable time throughout? (Replies came back in days, not weeks.)
Would work again. Would you take another deal from this brand? (Creators come back. The strongest single signal.)

The overall score

A brand’s overall reliability score is the mean of its four dimension averages, where each dimension average is taken across every published report. We round to one decimal place. Bands: 4.0–5.0 Reliable, 3.0–3.9 Mixed, below 3.0 Caution.

The 3-report minimum

No public score appears until a brand has 3 or more reports. This is the single most important rule in the product. It protects creator anonymity — a score built on one or two reports could be traced to an identifiable person — and it stops a single experience from defining a brand.

Anonymity

Reports are anonymous by default. We publish a creator type (such as “UGC creator”) and a banded deal value, never an identity or an exact figure that could deanonymise the creator. Aggregate-only display below the threshold reinforces this.

Moderation

Every report passes a moderation gate before it is published. Reports must reflect a real, completed deal the creator was part of, and must stick to verifiable facts — dates, terms, and outcomes — not insults. Reports that breach the guidelines are removed.

Brand right of reply

Brands can claim their profile and publish one response per report. A brand that owns a mistake and explains the fix often earns more trust than a brand with no track record at all. Right of reply keeps the registry fair to both sides.

Recency

Reliability changes — a finance-system migration or a new ops lead can move a brand in a quarter. We surface report dates prominently so readers can weight recent experiences appropriately.

FAQ

How is the overall reliability score calculated?: It is the mean of four equally-weighted dimensions — paid on time, brief clarity, responsiveness, and would-work-again — each rated 1 to 5 by creators after a paid deal, averaged across all of a brand’s published reports.
Why is a minimum of three reports required?: A single report can’t be a fair public verdict and could identify the creator. Requiring three reports protects anonymity and converts anecdote into a reliable signal.
How are fake or malicious reports handled?: Reports pass a moderation gate before publishing, must reflect a real deal, and stick to verifiable facts. Brands get a right of reply, and reports that violate the guidelines are removed.