The user rating summarises the numerical scores in eligible reader reviews published on a company’s review page. It is separate from the editorial expert rating.
On this page
What readers rate
Company reviews use three criteria, each scored from 1 to 5:
- Terms of Cooperation: the reviewer’s experience of the terms, fees and conditions of using the service.
- Platform Stability: the reviewer’s experience of the platform’s reliability and operation.
- Fulfillment of Financial Obligations: the reviewer’s experience of the company meeting its financial commitments.
These scores express the reviewer’s experience and opinion. They are not an independent audit of the company.
How an individual review score is calculated
We take the arithmetic mean of the valid numerical criteria in that review. Each included criterion has equal weight.
For example, scores of 1, 3 and 1 give (1 + 3 + 1) ÷ 3 = 1.666…, displayed as 1.7/5.
Can’t evaluate means that the criterion does not contribute a number. It is not treated as zero or as a neutral score. For example, scores of 5 and 3 with one criterion marked Can’t evaluate give (5 + 3) ÷ 2 = 4/5.
If there are no valid numerical criteria, the review has no numerical score. A published, unflagged review can still be included in the review count.
How the company’s user rating is calculated
The company’s overall user rating is the arithmetic mean of the individual scores of its eligible rated reviews. Each rated review has equal weight, including a review with only one or two valid criteria.
The calculation uses the underlying individual averages before display rounding. The overall result is displayed to one decimal place.
For illustration, one review with scores 1, 3 and 1 and another with scores 5 and 3 give ((1 + 3 + 1) ÷ 3 + (5 + 3) ÷ 2) ÷ 2 = 2.833…, displayed as 2.8/5.
Where an average is displayed for a specific criterion, it uses only the eligible reviews with a valid numerical score for that criterion. Helpful votes do not change the rating.
Which reviews contribute
- The review must be published and approved.
- It must be a top-level review. Replies in a discussion do not contribute to the company’s rating or review count.
- It must not carry the Cheating Review label.
- To contribute to the numerical average, it must contain at least one valid score from 1 to 5.
The review count and rating count can therefore differ: an eligible review with no numerical score adds to the review count but not to the numerical average. If no eligible rated reviews remain, there is no calculated numerical user rating.
What the Cheating Review label means
This label flags concerns about a review’s authenticity. Possible indicators include duplicate text, messages associated with the same IP address, or signs of coordinated reputation manipulation, whether positive or negative.
The label is an editorial flag, not conclusive proof that a review is fabricated. A shared IP address alone does not establish that reviews are fake. The label can be applied by an editor; it does not mean that every review has passed an identity or transaction verification process.
A flagged review may remain visible, with dashes shown instead of its criterion scores. It is excluded from the company’s user rating, criterion averages and eligible review count.
The flag does not erase the stored scores. If it is removed, those scores can contribute again, provided the review is approved and otherwise eligible.
When ratings change
Publishing, approving, editing, removing or flagging a review can change the figures. A review awaiting approval, marked as spam or placed in the trash does not contribute. Reviews may be held for moderation before publication.
Investment blog articles use a separate, single article score rather than the three company-review criteria described above.
User rating and expert rating
The user rating reflects eligible reader scores. The expert rating is a separate editorial assessment and is not calculated from those scores. Read how we assign the expert rating.
User ratings are a summary of reported experiences, not a guarantee of a company’s safety or the outcome of using its services.