What is a sentiment score?
A sentiment score is a number that summarises how positive or negative a piece of text is. Most systems use a scale from −1, very negative, through 0, neutral, to +1, very positive.
How sentiment scores are calculated
Lexicon-based systems add up the scores of positive and negative words, adjusted for negation, intensifiers and contrast, then scale the total into the −1 to +1 range. Machine learning systems output a probability for each class, which is turned into a score. Either way, the score is only as useful as your ability to see why it came out as it did.
A sentiment score example
“Fast setup and the team was lovely” might score +0.8. “It works, eventually” might score around 0. “Third time asking, still broken” might score −0.7. Across an account, the trend of these scores over weeks says more than any single one.
Net sentiment score
Net sentiment is the share of positive mentions minus the share of negative ones, ignoring neutral. If 55% of mentions are positive and 25% negative, net sentiment is +30. It is easy to report but hides intensity, so pair it with the average score.
Sentiment score vs CSAT
CSAT asks customers to rate their satisfaction directly, usually after an interaction, and only some reply. A sentiment score is inferred from what every customer wrote, whether or not they answered a survey. CSAT tells you what people say when asked; sentiment tells you what they said unprompted. Customer Support Sentiment Analyzer shows the exact terms behind each score, so a rating can be checked rather than taken on trust.
Last reviewed 2026-11-01.