Reading Between the Lines: How to Interpret Online Reviews
Review volume and star rating are only the beginning. Here's how our AI reads the real signal in your review profile.
Online reputation (15% of the overall grade) is one of the most nuanced dimensions to score. A raw star average is almost meaningless without context. Our AI goes deeper.
What we actually read
We search across Google Reviews, Trustpilot, Yelp, BBB complaint history, Reddit, Glassdoor, and industry-specific platforms. The goal isn't to find the average star rating — it's to understand the *nature* of the complaints and the *pattern* of positive reviews.
Sentiment composition matters more than stars. A business with 4.2 stars where 90% of reviews are 5-star and 10% are 1-star (bimodal) is very different from one with 4.2 stars spread evenly across 3–5 stars. The bimodal pattern often means the business is polarizing — great for some customers, terrible for others. The why matters.
Recency and volume trends. A business with 500 reviews averaging 4.6 stars *five years ago* and 20 reviews in the past year averaging 3.1 is in a different position than one with consistent growth. Reputation decay is a real signal.
Nature of complaints. Scam complaints, fraud allegations, undelivered products, and data breaches are categorically different from "the service was slower than expected" or "pricing is too high." Our AI distinguishes severity.
Owner response patterns. Do they respond to negative reviews? Do they respond defensively or constructively? A business that replies thoughtfully to criticism gets credit; one that ignores or threatens reviewers does not.
Reddit and forum mentions. Reddit is the most honest review platform because it's harder to game. A search for "[brand] reddit" often surfaces experiences that would never appear on a managed review platform.
The review manipulation problem
Our AI is specifically trained to flag suspicious review patterns:
- Large volumes of 5-star reviews with identical phrasing
- Review spikes after long gaps in activity
- Reviews from accounts with no other history
- Reviews that read as promotional copy rather than personal experience
A business with 300 reviews that appear organic scores better than one with 3,000 that pattern-match to a purchased review campaign.
What a strong reputation score looks like
High-scoring businesses have: consistent review volume over time, a realistic distribution of stars (no business is 100% five-star), genuine negative reviews that are responded to well, active Reddit and forum presence without major controversy, and BBB accreditation with zero unresolved complaints.
Improving your reputation score
The fastest way to improve reputation is to make it easier for happy customers to leave reviews. Most dissatisfied customers review without being asked; most satisfied ones don't.
A simple post-service email asking for a Google review — sent within 48 hours of service completion — can double review volume within six months. Genuine volume from real customers dilutes the impact of bad-faith negative reviews and creates a more accurate picture.
See how your site scores on Online Reputation
Full audit — all 10 dimensions, cited sources, and a shareable report.
Start a scan →