Our Review Methodology

Choosing the right software isn't always an easy task, especially when relying on online reviews. Some reviewers are notoriously harsh, refusing to give perfect scores. Others might leave one-star ratings with no explanation, while many hand out five stars without much thought or detail. That's why we've created the TrustScore - a rating system designed to go beyond surface-level reviews. It captures the genuine thoughts, feelings, and experiences of users, offering a clearer, more reliable picture of the software you're considering.

By relying on traditional review sites, you risk making important decisions based on incomplete, manipulated, or outdated information. Because traditional review sites rely on star ratings and simple ratings, they provide a false sense of trust and transparency.

Fake Reviews Are Everywhere
It's becoming increasingly hard to separate genuine feedback from fabricated reviews. This manipulation undermines trust and leaves users questioning the validity of what they're reading.
No Verification of Reviewers
Many platforms don't verify who's leaving reviews, which means anyone, including competitors or bots, can post feedback, skewing the overall perception.
Outdated Feedback Can Mislead
Older reviews often stay at the top, even if they no longer reflect the current state of the product or service. A tool that's been significantly improved, or deteriorated, might not be represented accurately.
Oversimplified Ratings Lack Depth
Simple star ratings or "good/bad" labels don't capture the nuances of user experiences. A five-star rating doesn't explain why the product was great, and a one-star rating doesn't provide insights into specific shortcomings.
Reviews Can Be Influenced
Some platforms allow companies to pay for certain privileges, like suppressing negative reviews or promoting positive ones, which compromises transparency and trustworthiness.
Context Matters, but It's Missing
Without context or standardized interpretation, reviews can feel subjective and unhelpful. What one person considers a minor inconvenience, another might see as a dealbreaker, and star ratings alone can't convey those differences.

We think it's time for a change. It's time for a more transparent, verified, and objective approach to understanding customer feedback and reviews. That's why we have decided to do things differently and create a new metric called TrustScore.

Our Data Collection Process

Our TrustScore is based on "real" reviews, and we gather anonymized data monthly from leading and trusted online public review platforms, such as Trustpilot, G2, Capterra, and TrustRadius.

Trustpilot Logo
G2 Logo
Capterra Logo
TrustRadius Logo

Each review undergoes both a human review and an automated validation process to ensure accuracy and authenticity. This systematic approach provides a continuous stream of current and relevant information from diverse user experiences.

TrustScore Explained

We decided to go beyond simple star ratings. Our core metric is TrustScore, a scoring method we have developed that places less emphasis on star ratings and more focus on the genuine thoughts, feelings, and experiences of users.

Collect Reviews
We gather authentic user feedback and reviews from verified sources
Analyze Sentiment
We use advanced AI to understand the true meaning behind reviews
Calculate Metrics
We process numerical ratings and combine with sentiment analysis
Generate TrustScore
We create a final score that accurately reflects user satisfaction

Our TrustScore combines numerical star ratings with the power of advanced sentiment analysis powered by VADER. Our TrustScore rating system is calculated using the following formula:

TrustScore = (Sentiment × 0.6) + (Rating × 0.4)
Step 1
Normalization

To ensure both sentiment and star ratings contribute proportionally, we first normalize them to a 0-1 scale. Our sentiment analysis tool, VADER assigns scores ranging from -1 (extremely negative) to +1 (extremely positive). We convert this to a 0-1 scale. Traditional star ratings (1-5) are also converted to a 0-1 scale

Step 2
Weighted Combination

We then combine these normalized scores with a specific weighting. Sentiment analysis gets a 60% weight because we place a higher emphasis on the detailed written feedback, as it often provides richer context and nuanced insights. Star ratings gets a 40% weight because we still value the numerical rating as a quick indicator of overall satisfaction

Step 3
Final Conversion

Finally, the combined score is multiplied by 100 to provide an easy-to-understand TrustScore on a scale of 0-100

Example
Example TrustScore

A review that says "Great tool, saves me hours every day!" with a 4.5 rating would get:

Sentiment score of 0.8 (very positive text) × 0.6 weight = 0.48
Rating score of 0.9 (4.5 out of 5) × 0.4 weight = 0.36
0.48 + 0.36 = 0.84 × 100 = 84/100
Test Our TrustScore Calculator

Adjust the sliders to see how different ratings affect the TrustScore

Sentiment Score
0.5
Negative
Positive
Star Rating
3.0
1 star5 stars
54/100

Fair

Understanding TrustScore Ranges
90-100
Exceptional

Outstanding software with consistent positive feedback and high user satisfaction

75-89
Excellent

Very reliable software with strong positive sentiment and minor concerns

60-74
Good

Solid software with positive feedback but room for improvement

40-59
Fair

Mixed feedback with notable concerns but still functional

0-39
Poor

Significant issues reported with major concerns from users

While our methodology is robust, it's important to understand how to interpret TrustScores effectively. Here are some common misconceptions and tips:

  • Higher scores always mean better software (False): The "best" software depends on your specific needs.
  • All software categories should be compared directly (False): Different types of software have different expectations and scoring ranges.
  • Recent reviews matter more than older ones (Partially True): While recent reviews are valuable, a consistent pattern over time is also important.
  • A perfect score means perfect software (False): No software is without its minor flaws.

VADER Sentiment Analysis

Our "secret weapon" for understanding the true meaning of customer reviews is VADER (Valence Aware Dictionary and sEntiment Reasoner).

This tool analyzes text for emotional intensity, context, and emphasis, allowing us to go beyond simple positive or negative classifications. Learn more about VADER here.

VADER excels at capturing the nuances that traditional star ratings often miss, such as the intensity of emotion, contextual modifiers, punctuation emphasis, mixed sentiments, and even subtle indicators of sarcasm.

Here are some examples of how VADER interprets different types of feedback:

Example 1
Positive

"The software works well" - VADER Score: 0.4 (mildly positive)

Example 2
Positive

"The software is incredible! It saved us hours of work!!" - VADER Score: 0.9 (strongly positive)

Example 3
Positive

"Great features but the interface needs work" - VADER Score: 0.2 (slightly positive, capturing the overall positive leaning despite the criticism)

Example 4
Negative

"Constant crashes and poor support response" - VADER Score: -0.6 (moderately negative)

Example 5
Positive

"Despite the learning curve, once you understand the system, it's actually quite powerful" - VADER Score: 0.5 (moderately positive, recognizing the eventual positive outcome despite initial challenges)

Example 6
Negative

"While the features are impressive, the constant crashes make it unusable" - VADER Score: -0.3 (slightly negative, acknowledging good features but with overriding flaws)

Our Commitment to Data Security, Privacy, and Quality

We take data security and user privacy extremely seriously. Our infrastructure is SOC 2 Type II certified, ensuring robust security controls.

We employ end-to-end encryption for all review data and conduct regular security audits to maintain the highest standards of data protection.

We are also committed to compliance with regulations like GDPR and CCPA in our data handling practices. To further protect privacy, we employ anonymized review processing, adhere to strict data retention policies, and do not store personal information.

We also provide opt-out mechanisms for reviewers and implement regular data purging schedules. Our transparent privacy policy outlines these practices in detail.

Maintaining the quality and integrity of our data is paramount. We have a 72-hour correction window for verified inaccuracies reported by users. We conduct monthly data validation cycles and utilize automated fake review detection systems.

We also employ cross-reference checking across platforms and analyze historical trends to identify anomalies.

Our Commitment to Independence

Website Rating maintains complete independence from the companies we review. Our revenue is generated exclusively through affiliate links, never through payments from software companies or sponsored reviews.

This commitment to independence ensures that our TrustScores remain unbiased and trustworthy.

Our quality assurance processes, including automated validation, multiple data point verification, regular system updates, and continuous monitoring, further reinforce this commitment.

For questions or feedback, please contact us.