How Google Star Ratings Are Calculated

Google star ratings are calculated as a simple average: the sum of every published star rating divided by the number of reviews, rounded to one decimal place. There is no secret weighting for recent or detailed reviews in the average itself — but removed reviews, rounding, and an update lag of up to two weeks explain most cases where the number looks wrong.

Every restaurant owner has stared at their Google rating and done the arithmetic — and come up with a number that doesn't match what Google shows. Three new 5-star reviews arrived this week and the rating didn't budge. A single 1-star appeared and the score dropped more than feels fair. The restaurant across the street shows 4.5 with reviews that look worse than yours.

This guide is the full, honest explanation of how Google star ratings are calculated — aimed at restaurant owners in the Nordics and beyond. No speculation presented as fact: the maths, the edge cases Google actually documents, the myths that keep circulating, and the formulas you can use yourself to plan how many reviews you really need. For the strategy side — getting the reviews in the first place — our complete restaurant review strategy covers that separately.

The maths is simpler than you think

Google's own help documentation says the review score for a place is the average of all ratings published on Google for that business. In practice:

Overall rating = (sum of all star ratings) ÷ (total number of reviews), rounded to the nearest tenth.

Take a small restaurant with ten reviews: six 5-star, two 4-star, one 3-star, one 1-star. Add them up: (6 × 5) + (2 × 4) + (1 × 3) + (1 × 1) = 30 + 8 + 3 + 1 = 42. Divide by 10 reviews: 42 ÷ 10 = 4.2. Google displays 4.2.

That's the entire calculation. There is no second step that Google isn't telling you about. Many reputation-management companies sell "rating calculators", and they're useful tools — but all they implement is this same division, plus a guess at Google's rounding. The interesting part isn't the division; it's everything around it: what counts, what doesn't, and why the number on your screen sometimes disagrees with your spreadsheet.

Why your own maths may disagree with Google

Four documented or observable reasons explain almost every mismatch between your calculation and Google's display.

Rounding hides small movements

Google rounds to one decimal. If your true average moves from 4.31 to 4.34 after a new review, the display stays at 4.3 — nothing happened from the guest's perspective. You need the true average to cross a rounding boundary (4.35 → 4.4) before anyone sees a change. On a profile with hundreds of reviews, a single 5-star review moves the true average by a fraction of a point and the display often doesn't move at all.

The update lag is real

Google states that a new review can take up to two weeks to affect the published score. New reviews are verified for authenticity before they count, and the profile re-aggregates on Google's schedule, not the guest's. So if three 5-star reviews arrived on Monday and the rating hasn't moved by Friday — that is normal, not a bug. Wait the two weeks before worrying.

Removed reviews change the math silently

Reviews Google removes under its content policies — spam, conflicts of interest, reviews of a visit that never happened — drop out of the average entirely. You are not notified in a way that makes the maths obvious: the count falls and the average shifts without a visible event. If your rating drops overnight with no new review, check whether your review count also dropped. If it did, Google took something away.

You forgot a review

The humble one: owners often compute the average from the reviews they can see on the first page of their listing. Google's reviews tab paginates, and old or short reviews hide several pages deep. Make sure you're counting from the full list — or better, do the calculation in reverse from the displayed average and your known count, as described in the five-minute check below.

The word that does all the work: "published"

One word does all the work in Google's definition, and it's published. The review score is the average of all ratings published on Google for that place or business. A rating that never publishes — because Google's automated systems filtered it under the Maps user-generated content policy — was never in the average to begin with. It isn't subtracted later. It never counted.

This matters more than it sounds. Common reasons a review never publishes: it came from an account with no history that posted several reviews in a burst, it was written from the same IP address as the business (the classic mistake is handing guests your own tablet at the counter and asking them to review on the spot), or the text tripped a spam filter. You see the review in your notification; your rating doesn't move. From Google's perspective there is nothing to move — the review isn't in the pool.

The flip side: reviews that do publish but are later removed drop out cleanly, and the average is recomputed without them. Losing a published 5-star review hurts the average exactly as much as gaining one helped it — which is why a sudden unexplained rating drop usually means Google removed something, not that the formula changed.

What doesn't change the number (the weighting myths)

Search this topic online and you'll find confident claims that Google weights recent reviews more, that long detailed reviews count more than star-only ratings, or that Local Guide reviews carry extra influence. Here is the uncomfortable truth: nothing Google publishes says any of that is true for the average itself. As far as the documented calculation goes, every published review counts exactly the same — a 5-star from three years ago and a 5-star from yesterday have identical weight in the division.

A few related things that also do not change the number:

• Your replies. Responding to reviews does not alter the maths. It changes how guests read the rating — which can matter more for bookings than a tenth of a star — but the division is untouched.
• Photos attached to reviews. Great for trust and for ranking in the local results; irrelevant to the average.
• Owner-verified responses or posts. Your Google Business Profile activity influences visibility, not the score.
• Reviews from the web. The "reviews from the web" that Google sometimes shows on your listing (from Tripadvisor or Facebook, for example) are display-only. They are not part of the Google rating calculation.

Now the nuance, because it's important for your strategy: while recency doesn't change the average, it changes almost everything else. Guests read recent reviews far more than old ones — most people only pay attention to reviews from the last few months. Google's local ranking also considers review freshness and velocity alongside the rating. So "new reviews don't change the average much" and "new reviews matter enormously" are both true. The number is a simple average; the business value around it is anything but simple.

How many 5-star reviews you need to reach your target

This is where volume turns into inertia. The more reviews you already have, the more new 5-star reviews it takes to move the average — because every new review is diluted by the existing total. A restaurant with 20 reviews can jump from 4.2 to 4.5 with a good month; the same restaurant with 400 reviews barely moves with a good year.

The exact formula, assuming every new review is 5 stars (so treat the result as a best case, and add a 30–50% buffer for a realistic mix):

Reviews needed = current review count × (target rating − current rating) ÷ (5 − target rating)

Example: you're at 4.2 with 120 reviews and want 4.5. That's 120 × (4.5 − 4.2) ÷ (5 − 4.5) = 120 × 0.3 ÷ 0.5 = 72 five-star reviews. Notice the denominator: the closer your target gets to 5.0, the larger it gets — climbing the last tenths is dramatically more expensive than climbing the first ones.

Two practical takeaways. First, fix a low rating early: recovering from 4.0 with 50 reviews takes a month or two of steady work; recovering from 4.0 with 500 reviews is a year-long project. Second, the mix matters — if even a fifth of your new reviews are 4-stars instead of 5-stars, the numbers above grow. Plan with the buffer, not the raw formula.

How many 5-star reviews cancel out a 1-star review

A related question owners ask constantly — and the answer surprises most of them: the number depends only on the rating you're defending, not on how many reviews you have. The formula:

Five-star reviews needed = (your current rating − 1) ÷ (5 − your current rating)

At 4.5, that's (4.5 − 1) ÷ (5 − 4.5) = 3.5 ÷ 0.5 = 7 five-star reviews to offset one 1-star. At 4.8, it's (4.8 − 1) ÷ (5 − 4.8) = 3.8 ÷ 0.2 = 19. The higher your rating, the more expensive each new 1-star review becomes to absorb — which is the mathematical reason the best review strategy is prevention: one avoided bad experience is worth a dozen solicited good ones.

For restaurants this lands concretely. A single bad Saturday — a long wait, a cold dish, an overwhelmed kitchen — can produce three or four 1-star reviews in one evening. At a 4.6 rating, absorbing four 1-stars needs roughly 4 × (4.6 − 1) ÷ (5 − 4.6) = 4 × 9 = 36 five-star reviews. That's months of normal review velocity erased in one night. The maths is the strongest argument for operational consistency that a sceptical partner will ever see.

How Google differs from Tripadvisor and Facebook

Google's simple average is actually the odd one out. Most other platforms do weight things:

Why does this matter for a Nordic restaurant? Because your guests don't live on one platform. Tourists check Tripadvisor, locals check Google Maps, and Facebook groups still drive discovery in smaller cities. But Google is where the rating is most visible — it's baked into Search and Maps, and it's what the majority of diners check first. Prioritise Google for volume, keep Tripadvisor healthy for tourists, and don't let Facebook rot: an unanswered recommendation page still shows up in search results.

What actually moves your rating up

If the formula is just an average, the strategy is almost insultingly simple: add more 5-star reviews than anything else, and avoid 1-star reviews. The execution is where restaurants win or lose, and it comes down to three habits.

Steady velocity beats bursts

Ten reviews a month for a year moves the rating more reliably than 120 reviews in a single week. Bursts also trip Google's spam filters — a sudden spike of reviews from similar locations or new accounts gets filtered, which means those reviews never publish and never count. The math rewards the boring, consistent ask: a few genuine reviews every week, indefinitely.

Ask — but ask within the rules

Google explicitly allows businesses to ask for reviews with a direct link or QR code. What it prohibits is incentives (anything of value in exchange for a review), selective solicitation (only asking happy guests), and pressuring guests on your premises. The clean, durable method is the one we detail in how to get more Google reviews: a direct review link on receipts, a QR code on table cards, a follow-up message after delivery — plus staff who know the one sentence to say at the right moment. Getting your staff behind the review strategy is what turns the system from a poster on the wall into actual weekly volume.

Defend the floor, don't chase the ceiling

The formulas above show that climbing from 4.6 to 4.8 costs far more than climbing from 4.0 to 4.3. For most restaurants, the realistic target is a stable 4.3–4.6 with steady velocity, not a perfect 5.0 — which guests don't even trust. Chasing perfection also invites the behaviour that gets profiles penalised: review gating, incentives, asking staff to review the business. A 4.5 with 200 genuine reviews and thoughtful owner replies converts better than a suspicious 5.0 with 40.

Check your own maths in five minutes

Want to verify Google's number — or figure out exactly how many reviews you need? Here's the quick version:

1. Count your reviews from your Google Business Profile dashboard (not the public page — the dashboard shows the true published count).
2. Estimate your true average. If Google shows 4.3, your true average is somewhere between 4.25 and 4.34. For planning, use the middle: 4.295.
3. Pick a target — 4.5 is the practical sweet spot for most restaurants.
4. Apply the formula: reviews needed = count × (target − current) ÷ (5 − target), then add a 30–50% buffer for mixed ratings.
5. Check monthly: track the true average, not the display. You'll see progress in the decimals weeks before it appears as a new digit.

One caution about the free "Google review calculators" online: they're accurate only if your inputs are. A rounded 4.3 can hide very different true averages — 4.25 needs nearly twice as many new 5-stars to reach 4.5 as 4.34 does. Always enter your full average when the dashboard shows it, and treat any calculator's answer as a floor, not a forecast: it assumes every new review is 5 stars and that every review publishes.

Frequently asked questions