Anatomy of a Trust Score

The Composite Business Score is a number between 0 and 100. It appears simple. It is not.

Behind that number is a scoring engine that weighs reviewer credibility, adjusts for industry norms, applies time decay, detects manipulation patterns, and calibrates across 27 industry categories. The result is a score that is extremely difficult to game and genuinely reflective of customer experience.

Not an average

The simplest approach to scoring would be averaging all review scores. This is what most platforms do. It is also trivially exploitable: buy ten five-star reviews to overwhelm three genuine one-star reviews.

VeriBureau does not average. Each review is weighted by three factors: the reviewer's Trust Score, the age of the review, and the industry calibration. A review from a highly trusted reviewer with decades of verified interactions contributes more than a review from a new account. A recent review matters more than one from last year. And the thresholds for "good" vary by industry — a 70 in healthcare represents something different than a 70 in retail.

Time decay

A business that was excellent two years ago and terrible today should not coast on historical reputation. VeriBureau applies time-based decay to every review. Recent interactions weigh more. Old reviews fade gradually. The score always reflects current performance.

The decay rate is calibrated per industry. A restaurant changes faster than a telecommunications provider. The scoring engine accounts for this.

Anti-manipulation

The scoring parameters are proprietary. This is intentional. Publishing the exact weights and thresholds would be equivalent to publishing an exploitation manual. Every spam operation would immediately optimize for the published formula.

What we can say: the engine uses logarithmic scaling for review volume, which means the marginal impact of each additional review decreases. The first ten reviews have enormous impact. The next hundred have progressively less. This makes it economically unviable to buy enough reviews to significantly move a mature score.

Coordinated patterns — multiple reviews from low-trust accounts within a short time window — trigger automatic weight reduction. The more a manipulation attempt looks like a manipulation attempt, the less effect it has.

What the score means

Scores above 80 represent consistently excellent verified experiences. Scores between 60 and 80 indicate good performance with room for improvement. Scores between 40 and 60 suggest mixed results. Below 40 indicates significant issues.

These are not arbitrary thresholds. They are calibrated against industry-specific distributions and adjusted as the dataset grows. The goal is that a score of 75 in telecommunications means roughly the same thing as a score of 75 in hospitality: above average, reliably good, with occasional issues.

A living number

The Trust Score is not computed once. It is recomputed with every new review. It responds to trends — a sudden drop in review quality is reflected immediately, not averaged away. It rewards consistency and penalizes volatility.

This is not a rating. It is a continuously verified measurement of how well a business serves its customers, as reported by the customers themselves, weighted by their own proven credibility.