AI
Review sentiment analysis
Review sentiment analysis is the automated classification of review text by sentiment and theme, usually to surface recurring operational problems.
Star ratings alone are a poor operational signal because a 4-star review can contain a serious complaint and a 5-star can contain nothing actionable. Theme extraction across a few hundred reviews is where the pattern appears.
Modern language models make this cheap enough to run monthly on a small business's full review history, which was not true a few years ago.
In practice
Run it on the last twelve months, not the last week. You are looking for the complaint that appears eleven times, not the one that appeared loudest.
Common questions
- Do I need a tool for sentiment analysis?
- At under fifty reviews a month, reading them is faster. Above that, tooling stops being optional.