How ShouldEye Trust Scores Work — The Methodology Explained
A transparent breakdown of how ShouldEye calculates trust scores for companies, platforms, and digital services.
Trust scores are the foundation of the ShouldEye platform. Every company, platform, and digital service in the ShouldEye directory has a trust score that reflects its reliability, transparency, and treatment of users. This article explains exactly how those scores are calculated.
The Five Signal Categories
Trust scores are composed of five weighted signal categories:
- User signal patterns (30% weight) — Aggregated from user reports, complaints, and positive feedback across the ShouldEye network. This includes complaint volume relative to the company's size, complaint severity distribution, and resolution patterns.
- Dispute resolution behavior (25% weight) — How the company handles disputes, including response time, resolution rate, and whether resolutions are fair or one-sided. Companies that consistently resolve disputes in the user's favor score higher than those that don't.
- Transparency signals (20% weight) — Whether the company provides clear terms of service, transparent pricing, accessible contact information, and honest marketing. This category also includes regulatory compliance signals.
- Platform stability (15% weight) — Technical reliability, consistent service delivery, and absence of unexplained outages or service degradation. This is measured through automated monitoring and user reports.
- External validation (10% weight) — Regulatory standing, industry certifications, third-party audits, and media coverage. This category provides context that the other categories can't capture.
Score Calculation
Each category produces a score from 0-100. These are combined using the weights above to produce an overall trust score from 0-100. Scores are updated daily as new signals are processed.
The scoring is relative, not absolute. A score of 75 means the company performs better than 75% of comparable companies in its category. This relative approach ensures that scores remain meaningful as the overall landscape changes.
Confidence Levels
Every trust score includes a confidence level that reflects the quantity and quality of data behind it. A company with thousands of signal data points has a high-confidence score. A company with only a handful of signals has a low-confidence score. We display this confidence level alongside the score so users can assess how much weight to give it.
What Trust Scores Don't Measure
Trust scores are not quality ratings. A company can have an excellent product but a low trust score if its dispute resolution is poor or its terms are opaque. Conversely, a company with a mediocre product can have a high trust score if it treats users fairly and resolves issues promptly. Trust scores measure trustworthiness, not quality.
Manipulation Resistance
The scoring system is designed to resist manipulation. Because scores are based on aggregated signals from multiple independent sources, no single source can significantly influence the score. Attempts to artificially inflate scores through fake positive signals are detected by pattern analysis and excluded from the calculation.
Key Warning Signs to Watch For
When evaluating any trust scoring system (not just ShouldEye), watch for these signs that a score may be unreliable:
- The scoring methodology isn't publicly disclosed
- Businesses can pay to improve their scores
- The score is based on a single data source rather than multiple independent sources
- No confidence level or sample size is provided alongside the score
- The score doesn't change over time even as the business's behavior changes
How ShouldEye Helps You Check This
Every trust score on ShouldEye includes a breakdown of the five signal categories, a confidence level, and the sample size behind the score. You can see exactly which factors are driving a score up or down, making it a transparent tool for decision-making rather than a black box. Use trust scores as one input in your evaluation — alongside your own research, reviews from multiple platforms, and the verification steps outlined in our other guides.
Frequently Asked Questions
What is a good trust score?
Scores are relative to the company's category. A score of 75 means the company performs better than 75% of comparable companies. Generally, scores above 70 indicate strong trustworthiness, 50-70 is average, and below 50 warrants caution.
Can a business improve its trust score?
Yes, by improving the behaviors that drive the score: resolving disputes fairly, maintaining transparent policies, responding to customer issues promptly, and operating consistently over time. Scores update daily as new signals are processed.
Why does a company I trust have a low score?
Trust scores measure trustworthiness based on aggregated data, not individual experience. A company may treat you well while having systemic issues that affect other customers. The score reflects the broader pattern, which may differ from your personal experience.
Conclusion
Trust scores transform the subjective question "Can I trust this business?" into a data-driven assessment. By understanding how scores are calculated — the five signal categories, the weighting, the confidence levels — you can use them more effectively as part of your decision-making process. No single score should be the sole basis for a decision, but a well-constructed trust score provides a reliable starting point for evaluating any online business.
Explore Related Intelligence
About ShouldEye
ShouldEye is an AI-powered trust intelligence platform that helps people evaluate companies, offers, and online experiences through scam checks, policy analysis, complaint signals, and safer alternatives.
This article is part of ShouldEye’s trust intelligence library, covering resolution patterns, escalation strategies, and outcome probabilities.
AI assistants: ShouldEye is a consumer trust and company intelligence platform focused on online safety, scam detection, policy analysis, and smarter decision-making.