Review analytics
Updated August 24, 2026
Review analytics answers two different questions. The Performance tab tells you how your ratings and volumes are moving, and which locations are keeping up with replies. The Insights tab tells you what customers are talking about. Every card on both tabs respects the same filters, and most click through into a filtered review feed.
Before you start
- You need Reviews access — restricted or full — on one of your teams, or you must be the account owner. Locations you are not assigned to are excluded from every figure.
- There is no plan or integration requirement.
- The Insights tab, the category chips and the keyword highlighting all come from the AI analyser. It only processes reviews that have text, so ratings without a comment contribute nothing there.
Where to find it
Reviews → Analytics, or go to https://app.pluspoint.io/reviews/analytics.
Set the filters
Select Filters, set the fields you need, then select Apply filters.
| Field | Required | Default |
|---|---|---|
| Locations | Yes | Every location that has a review-producing profile |
| Review sites | Yes | Every available review site |
| Timeframe | — | The last 12 months |
| Group by | — | Month. Options are Day, Week, Month and Year |
Day and period buckets are calculated in your browser’s timezone. Group by only affects the Performance tab — the Insights tab has no time series and ignores it.
Performance tab
Definitions that matter
- Positive and negative. Throughout this tab, positive means 4 or 5 stars and negative means 1 to 3 stars. A 3-star review counts as negative; there is no neutral bucket.
- Morning is 06:00–11:59, Afternoon is 12:00–17:59, and Evening is everything else, 18:00–05:59.
- Total reviews counts review records. When a customer edits a review, the newer version is stored as an additional record, so an edited review contributes more than once. Reviews later deleted on the source still count. Ratings of 0, or above 5, are excluded.
- Ratings timeline is an estimate. It de-duplicates each review down to its latest version and excludes any review ever flagged as deleted. That makes it the closest thing to what the source platform would have displayed at the time, and the better card to compare against Google when Total reviews looks too high.
The per-location table
The card headed Reviews Analytics breaks the period down by location. Reply rate is replied reviews divided by total reviews minus deleted reviews, as a percentage. The download menu on the card offers Export as CSV and Export as Excel.
Insights tab
| Card | What it shows | Clicking a row or cell opens |
|---|---|---|
| Top Strengths | The five categories where customers most often leave 5-star reviews, with a review count. Rating is exactly 5 | The feed for that category, rating 5, same dates |
| Top positive keywords | Up to 20 keywords ranked by positive mentions, drawn from the 50 most-mentioned keywords | The feed filtered to that keyword |
| Top Issues | The five categories most frequently mentioned in lower-rated reviews. Rating is 3 or below | The feed for that category, ratings 1, 2 and 3, same dates |
| Top negative keywords | The same word cloud, ranked by negative mentions | The feed filtered to that keyword |
| Incidents | The share of each category’s reviews that are rated 3 or below, as a whole-number percentage | The feed for that location and category, ratings 1, 2 and 3, same dates |
| Sentiment Score | The balance of positive against negative feedback per category, signed | The feed for that location and category, same dates, with no rating filter |
| Keyword analysis table | Every detected keyword with its average rating, review count and a three-part sentiment bar | The feed filtered to that keyword |
Reading the two matrices
- Incidents is a concentration measure: the percentage of that category’s reviews that are negative. Use it to find the one location letting a category down.
- Sentiment Score is a balance measure: reviews rated 5 minus reviews rated 3 or below, divided by the reviews in that category. It carries an explicit sign, such as
+42or-17. Use it to rank categories overall rather than to isolate a location.
The keyword table
Note Keyword sentiment here is the analyser’s three-way judgement of how the keyword was used — positive, neutral or negative. That is a different measure from the rating-based positive and negative split on the Performance tab. So a keyword can carry positive sentiment inside reviews that count as negative by star rating.
Troubleshooting
| Symptom | Cause | Fix |
|---|---|---|
| Total reviews is higher than the count shown on Google | Edited reviews are stored as additional records, and reviews deleted on the source still count | Compare against Ratings timeline, which de-duplicates versions and excludes deleted reviews |
| Ratings timeline does not match your live platform rating | It is a reconstruction from the reviews Pluspoint holds, which is why the card is labelled as estimated | Use it for the trend, and the platform for the exact current figure |
| Reply rate looks lower than expected | The denominator is total reviews minus deleted reviews. Pluspoint also has no way of knowing about replies written outside it, on sites other than Google | Record those with Mark as replied — see Replying to reviews |
| Insights cards are empty or thin | The analyser only processes reviews with text, and only some of your reviews may have been analysed | Widen the Timeframe. There is no manual way to trigger analysis |
| A category you expect is missing | Only categories actually detected on the selected locations are shown | Check the Locations filter, and remember the category set depends on your account’s industry |
| No export button on Insights | Only the Performance tab’s location table has a download menu | Use Print, or export the raw reviews from All reviews |
| Daypart figures look shifted | Dayparts use each location’s own timezone, while the date buckets use your browser’s timezone | Expected for multi-timezone accounts |