Two different measures
| Aspect | Star rating | Text sentiment |
|---|---|---|
| Measures | Overall verdict on a five-point scale | How customers feel about each topic |
| Use it for | Headline performance and what customers see | Finding what to fix |
| Comparing stores | Context; dominated by volume and history | Current, specific differences within peer groups |
| Early warning | Can rise while a topic's complaints grow | Topics often move first |
| In reports | At the top, with count and recent window | Beneath: topics that moved, with counts and quotes |
The question of sentiment vs star rating is really about what each one measures. A rating is the customer's overall verdict on a five-point scale. Sentiment from text describes how they felt about each thing they mentioned. Neither is wrong; they answer different questions.
When they disagree
- Five stars with a complaint: "Loved it, but parking was terrible."
- Three stars with mostly praise: a middle rating for one problem.
- One star with praise: anger about one thing, appreciation for another.
- Rating-only reviews: a verdict with no explanation.
These disagreements are where the useful detail is. A complaint inside a five-star review is often the most fixable feedback you get.
Which to use for what
- Headline performance and what customers see: the rating, recent and all-time.
- Finding what to fix: topic sentiment from text.
- Comparing stores: topic negative shares within peer groups, with ratings for context.
- Checking a fix: the topic's negative share before and after.
Reporting both
Put the rating at the top with its review count and recent-window figure. Beneath it, show the topics that moved, with negative share, counts and quotes. When the rating moves, the topics show why; when topics move first, they warn you.
Star rating vs review sentiment: why the average lags.
Star rating vs review text why the average lagsFree review snapshot
Want this done for your own reviews?
No commitment. We use your details only to send your snapshot and follow up. Privacy policy
An example
A store's rating rises slightly over a quarter, and leadership is pleased. Text sentiment shows staff praise up strongly, while checkout complaints doubled, mostly in four- and five-star reviews. Without the text, the checkout problem would have stayed invisible until it pulled the rating down.
Common mistakes
- Trusting the rating alone.
- Replacing the rating with an invented sentiment index.
- Ignoring complaints in positive reviews.
- Reporting either without counts.
Common questions
Which is more accurate?
Each is accurate for what it measures; they measure different things.
Should sentiment ever override the rating?
For deciding what to fix, yes; for headline performance, report both.
Do customers see sentiment?
They see ratings and read text; Google also shows summaries and snippets.
Can we predict ratings from sentiment?
Topic sentiment usually explains rating changes; prediction is less important than explanation.
Checklist
- Rating as headline with count and recent window.
- Topic sentiment beneath it.
- Complaints in positive reviews captured.
- Counts beside every number.
In short
Ratings show the overall verdict customers see; topic sentiment shows what drives it. Report both, and use the text to decide what to fix.
For store comparisons
When comparing stores, ratings are dominated by volume and history, while topic sentiment shows current, specific differences. Use the rating to see which stores customers rate lower, and topic sentiment to see why, within peer groups and recent windows.
Explaining it to leadership
Show both side by side for one store over a year: the rating line barely moving, and two or three topic lines moving clearly. One example like this usually makes the case for topic sentiment better than any explanation.
A quick test for your own data
Take a month of reviews, find every four- and five-star review that contains a complaint, and count them by topic. If one topic stands out, you have found something your rating report has been hiding.
Doing this with Falcon
Falcon measures sentiment for every theme a review mentions rather than one score per review, keeps the exact words that support each label, and reports negative share by topic, store and month with small samples flagged and unclear texts kept visible. Every number opens the reviews behind it.