Sentiment & trends
Review sentiment & trends
See the mix of positive, neutral and negative reviews across your App Store and Google Play apps for the last 7, 30 or 90 days or the past year. Today sentiment comes from star ratings, and we say so plainly.
14-day free trial · cancel anytime
How sentiment is mapped today
A fixed rule based on the star rating of each review. No model, no black box.
Positive: 4–5 stars
Users who are broadly happy, even if they mention a small issue.
Neutral: 3 stars
Mixed reviews. Often the most useful to read: they like it, but something is in the way.
Negative: 1–2 stars
Bugs, crashes, billing problems and missing features usually land here.
What you can do with app review sentiment analysis today
The dashboard shows the positive, neutral and negative breakdown for the time range you pick (last 7 days, 30 days, 90 days or year) and the apps you select, next to the full rating distribution, reviews over time and reviews by platform.
- Spot a bad release. Switch between the last 7 and 30 days. If the negative share is much higher for the most recent week after a version ships, filter the review table to 1–2 stars and that platform, and read what changed.
- See movement the average hides. More 1-star and more 5-star reviews at the same time can leave the average flat. The split shows both.
- Compare iOS and Android. Select one app at a time, or use the platform filter on the review table, to see whether the negative trend is on one store or both.
- Route negative reviews. Give a destination a maximum rating of 2 and only negative reviews reach that Slack channel or Linear team.
Because review history is imported when you connect an app (including up to 36 months of Google Play reviews older than the API's one-week window), you can compare today with previous releases from day one.
The limits of rating-based sentiment
Star ratings are a reasonable proxy for how a user feels overall, and they are transparent: you always know why a review was counted as negative. But they miss things, and you should know where:
- Mixed reviews. “Five stars, but the latest update logs me out every day” is counted as positive, even though it reports a bug.
- Feature requests. A 2-star “please add offline mode” is counted as negative even if the tone is friendly.
- No topics. Rating-based sentiment tells you how many users are unhappy, not what about. For that you still read the reviews, or route them by keyword.
- Star-only ratings. Ratings without a written review are not available as reviews through the store APIs, so they are not in the breakdown.
Until text-based analysis ships, the practical workaround is keyword filters on destinations: send reviews that mention “crash”, “login” or “refund” to the right team, whatever their rating.
Today vs coming soon
| Today | Coming soon | |
|---|---|---|
| Sentiment source | Star rating (4–5 / 3 / 1–2) | Review text (NLP) |
| Breakdown by time range | ✓ | ✓ |
| Filter by app | ✓ | ✓ |
| Mixed-sentiment detection | ✗ | ✓ |
| Topics and keywords | Keyword filters on destinations | Automatic extraction |
NLP sentiment and topics are coming soon
We are working on sentiment and topic analysis that reads the review text. Want to try it first? Join the waitlist. Agents will also be able to query sentiment trends through the planned MCP server.
Frequently asked questions
From the star rating. Reviews with 4 or 5 stars count as positive, 3 stars as neutral and 1 or 2 stars as negative. The review text is not analyzed for sentiment today.
No. It is a simple, transparent rule based on rating. Natural-language sentiment that reads the review text is coming soon, and we will label it clearly when it ships.
When the stars and the words disagree. A 5-star review that says "great app but it crashes on login" is counted as positive, and a 2-star review that says "love it, please add dark mode" is counted as negative. Mixed reviews are common, so read the text before acting on a trend.
The average hides shifts. If 1-star and 5-star reviews both grow, the average can stay flat while more users are unhappy. A positive / neutral / negative split over time shows that movement.
Yes. Set a maximum rating of 2 on a Slack, Linear, Microsoft Teams, Zendesk or webhook destination and only negative reviews are sent there. Destination availability depends on your plan.
Email hello@review.expert and we will let you know when text-based sentiment and topic analysis are ready to try.
Track review sentiment across both stores
Connect your apps and see the positive, neutral and negative mix of your reviews, backed by your imported review history. 14-day free trial.