What sentiment analysis actually measures
At its simplest, sentiment analysis sorts text into positive, neutral or negative. Applied to a batch of reviews, that produces one useful but limited number: "63% negative." It's a real signal — but on its own, it doesn't say what's driving it.
Sentiment percentages
Review X-Ray scores every review on a genuine three-point scale — not a binary "good or bad" forced into three buckets — and reports the overall split alongside a risk read (low, moderate, or high friction) so the headline number has context.
Recurring topics
The next layer is knowing what customers are negative about. Sentiment is broken down by category — pricing, support, quality, shipping — so "63% negative" becomes "63% negative, and half of that is concentrated in billing."
Negative sentiment by topic
This is where a competitor weakness becomes actionable: not just that a category skews negative, but how many reviews, how severely, and what customers say about it directly.
Why sentiment alone isn't enough
Two products can both be "60% negative" and be in completely different situations — one facing widespread mild annoyance, the other facing a smaller number of customers furious enough to leave. Sentiment alone can't tell them apart. Severity can.
Sentiment tells you how customers feel. Review X-Ray goes further to tell you why — and how badly.
How Review X-Ray scores severity
Negative reviews get a second pass: minor annoyance, significant frustration, or dealbreaker — reserved for reviews where the customer says they stopped using the product, want a refund, or are actively warning others away. That distinction is what separates "worth monitoring" from "worth fixing first."