Customer review sentiment and theme analyzer
Paste a batch of reviews for your business or a competitor. The analyzer scores each review, reports the positive and negative share, groups comments into ten themes and lists the phrases customers repeat.
How to use it
- Copy 20-50 recent reviews from one platform (without reviewer names).
- Paste them with one review per line or a blank line between reviews.
- Read the sentiment index and the theme table.
- Compare with a competitor's reviews to find their recurring complaints.
Method
Each review is scored with a sentiment lexicon of positive and negative words, with negation handling (“not good” counts as negative) and intensifiers (“very”). Scores are normalised by review length. Reviews above +0.25 are positive and below −0.25 negative. Themes are detected with keyword groups for price, quality, service, speed and delivery, ease of use, reliability, features, ambience, location and refunds. The index maps the average score to 0-100, where 50 is neutral.
Worked example
Eight café reviews produce a sentiment index in the high 50s: praise for coffee and ambience, complaints about service speed and delivery temperature. “Customer service” appears as mixed, “Speed & delivery” as a complaint.
How to read the result
Look for themes that are complaints across several competitors (an unmet need) or for one strong rival (a switching opportunity). Differences of a few index points are noise.
Limitations
- Lexicon scoring misses sarcasm and some slang or local-language words.
- Review samples skew towards very happy and very unhappy customers.
- Read a sample yourself before acting on the numbers.