Google’s review score is the average of the ratings published for a business. To raise that average, the ratings entering the calculation must improve it. Posting replies or collecting newer reviews does not, by itself, change the arithmetic. Google explains how review scores are calculated.
Start with the numbers
Consider a hypothetical business with 40 ratings totaling 160 stars: its exact average is 4.0. If its next ten ratings are all five-star, the calculation becomes 210 stars divided by 50 ratings, or 4.2.
| Additional ratings in this example | Total ratings | Exact average | | ---------------------------------- | ------------- | ------------- | | None | 40 | 4.0 | | Ten five-star ratings | 50 | 4.2 | | Forty five-star ratings | 80 | 4.5 |
This is an illustration, not a target for staff to obtain five-star reviews. Actual outcomes depend on the full mix of ratings. A displayed rounded score is also insufficient to reconstruct the exact underlying total.
A new rating below the existing average lowers it; one above the average raises it. Changes to the underlying set of published ratings can also change the result.
Separate rating, freshness and search position
These measures answer different questions:
- Star rating: What is the average customer assessment?
- Review recency: How current is the feedback people can read?
- Maps position: Where does Google show the business for a particular search?
Use the review recency guide to assess freshness. For visibility, read the separate guide to Google Maps ranking for restaurants. An improved average cannot guarantee a particular search position.
Use the comments to choose a service improvement
Read a sample of recent reviews across the rating range. Group repeated observations into specific issues: an unexplained wait, inconsistent opening hours, an excellent welcome, or a confusing bill.
Choose one issue the team can act on. For example, if guests repeatedly mention waiting without an update, introduce a clear arrival-time explanation and assign someone to communicate delays. Check subsequent feedback for that theme rather than assuming every rating change came from the intervention.
Sentiment insights can help organize themes. They support the review process; they do not establish what caused a rating change.
Invite a representative range of experiences
Offer the same optional invitation to customers regardless of whether they complimented the service or raised a concern. Do not offer rewards or screen customers by expected rating. Google’s review policies explain the restrictions.
A suitable invitation is: “You’re welcome to share an honest review of your visit. The link is on your receipt.”
For a repeatable process, use the guide to collecting Google reviews. Keep service recovery available to everyone, independently of public posting.
What should you do with a negative review?
Investigate the issue, acknowledge what you can substantiate, and explain a practical next step. Avoid sharing private customer information. A useful response addresses the experience without pressuring the author to change their score.
If your team drafts responses with Starise’s response tools, review the details before publishing. An owner reply is a public explanation, not an extra rating.
FAQs
Can replying to reviews increase the average?
The reply itself adds no customer stars. It can explain your response to a problem, while the numerical average depends on the published ratings.
How many reviews do we need to reach a target rating?
There is no universal number. It depends on your current ratings and the scores of future reviews. The table above deliberately uses exact hypothetical totals to show that dependency.
Should we ask only happy customers?
No. Make an honest review equally available after all genuine experiences, with no reward tied to posting.
