How a star rating is actually calculated
The arithmetic is unglamorous, and it is the reason most advice about fixing a rating is wrong.
No credit card required. 30-day free trial.
Star rating, in one sentence
A star rating is the summary figure a platform prints beside a listing, computed from the individual ratings it has collected and shown to one decimal place.
The model to reason with is the plain average: add the ratings up, divide by how many there are. It is an average of all time, which is the single most important thing about it. A rating is not a measure of how you are doing. It is a measure of how you have ever done, weighted by how many people said so.
Not every platform prints a plain average, and two that GetSetReply reads say so themselves. Trustpilot states that a business's TrustScore is based on the number of reviews, how old they are, and a weighting system that includes seven reviews at 3.5 stars, so age carries weight and a new listing does not start from nothing. Yelp states that reviews its recommendation software does not recommend stay reachable from a link at the foot of the page but do not count towards the overall star rating or the review count, so some ratings are outside the figure entirely.
This is why two businesses showing the same rating can be in completely different situations, and why the rating on its own is a poor dashboard. Read it beside your velocity and your recent reviews, never alone, and check what the platform in front of you says it is computing before predicting what one review will do to it.
What one more review actually does to your average
This is the plain average, which is the right model unless the platform publishes another. Start with a listing holding 180 reviews at an average of 4.6, and add a single one star review to it. Then do the same to a listing with 20 reviews at the same average.
- Large listing before
- 180 reviews, 828 stars, average 4.6
- Large listing after one 1 star
- 181 reviews, 829 stars, average 4.58
- Small listing before
- 20 reviews, 92 stars, average 4.6
- Small listing after one 1 star
- 21 reviews, 93 stars, average 4.43
On the large listing the bad review moved the displayed number by nothing at all, because 4.58 still prints as 4.6. On the small one it moved 4.6 to 4.4. Volume is the shock absorber, and it is the only one there is.
Run it the other way and the lesson gets sharper. Lifting that same 180 review listing from 4.60 up to 4.70 takes 60 consecutive five star reviews with not one other rating in between, which at nine reviews a month is close to seven months. Below about fifty reviews the average is worth managing. Above a few hundred, your velocity and your replies are where the effort belongs.
Why the rating is a slow variable
The rating is the first thing a stranger sees, and it is compared rather than read. Nobody knows whether 4.4 is good. They know whether it is higher or lower than the place two doors down, and that comparison takes about a second.
The practical consequence is that fixing your rating this month is not a plan. Asking every customer for the next six months is a plan, and the rating is what it eventually produces. Treat the number as an output, never as a target you can attack directly.
It also means a young listing is fragile in a way an established one is not. Under twenty reviews, one unhappy Saturday is visible on your profile for the best part of a year.
The misconception: a 5.0 is the target
Plenty of businesses chase a perfect five, and some of them buy it.
A flawless average across a large listing removes the one thing that makes the good reviews believable, which is the presence of a few that are not. A reader who scrolls a hundred identical five stars starts looking for the catch.
There is a harder reason too. On a busy business the route to a perfect average normally runs through gating or through purchased reviews, and both are prohibited. Google's content policy forbids selectively soliciting positive reviews, and the FTC's 2024 rule bans buying and selling fake ones.
Questions about star ratings and averages
The arithmetic questions that come up most, answered plainly.
How is a star rating calculated?
Add every rating together and divide by how many there are, then round to one decimal place. That is the model to reason with, and two platforms say they do something else: Trustpilot weights by how old a review is and starts every business off with seven reviews at 3.5 stars, and Yelp keeps reviews its recommendation software does not recommend out of both the displayed rating and the review count.
Why did my rating not change after a new review?
Because the display rounds to one decimal. On a listing with a few hundred reviews a single new rating usually moves the true average by less than half of one tenth, so the printed number sits still.
Can I remove a bad review to fix my rating?
Only if it breaks the platform's own content policy, and you have to ask the platform rather than remove it yourself. Using false accusations, intimidation or a groundless legal threat to get a review taken down is the review suppression the FTC's 2024 rule prohibits.
Do star ratings from different sites combine?
No. Each platform computes and displays its own from its own reviews. A business commonly carries several ratings at once, which is why they are worth reading together rather than quoting whichever one flatters you.
Is a 4.7 better than a 5.0?
For a listing with real volume, many readers find a high but imperfect rating easier to believe. The safer framing is that a rating built by asking everybody is worth more than a higher one built by asking selectively.
The numbers that sit either side of your star rating
Move the input, not the output
Volume is the only lever on an average
GetSetReply hands you a funnel link and a QR code to collect with, gathers what comes back into one inbox, and drafts the reply for you to post.