Count the basis
Exact review count, exact sum behind the average and the list of entries that break policy. Without those three numbers any target is a guess, so we start there and show the arithmetic.
RATING RECOVERY · THE ARITHMETIC
From 3.5 to 4.6
is arithmetic,
not magic.
A Google rating is a sum divided by a count. Both sides of that fraction move: appeals take out reviews that break policy, and new reviews come from customers who were actually served. Here is what each path costs, in reviews, before you pay anyone for it.
01ADDING REVIEWS
What each target
costs in five-star reviews.
The formula is the whole trick: k = (T − R) ÷ (5 − T) × N, where T is the target rating, R is the current one and N is the number of reviews you have today. It assumes every new review is a five — which is the best case, and the case vendors love to quote.
| Target rating | Five-star reviews needed | Total reviews after | What it feels like |
|---|---|---|---|
| 3.8 | 30 | 150 | One good quarter of new reviews |
| 4.0 | 60 | 180 | Half a year at a steady flow |
| 4.3 | 137 | 257 | a year, and the flow cannot stall |
| 4.6 | 330 | 450 | 2.75x the profile you have now |
| 4.8 | 780 | 900 | a different business, not a campaign |
Read the 4.6 row again: 330 published five-star reviews, not 330 requested. If a vendor promises that jump in a month on a profile with 120 reviews, the only way to deliver it is to invent the reviewers.
02REMOVING REVIEWS
Appeals move the number
faster — where they apply.
Removing a one-star entry changes both parts of the fraction at once: the sum drops by one point and the count drops by one review. That is why fifteen successful appeals do more than a hundred new requests.
| One-star reviews removed | New rating | Reviews left | Cost |
|---|---|---|---|
| 5 | 3.61 | 115 | 5 evidence packs |
| 10 | 3.73 | 110 | 10 evidence packs |
| 15 | 3.86 | 105 | 15 evidence packs |
| 20 | 4.00 | 100 | 20 evidence packs |
| 25 | 4.16 | 95 | 25 evidence packs |
Where this stops: appeals only touch reviews that break policy. A real customer who had a bad experience stays on the profile, and no form removes them. That is the difference between reputation recovery and review laundering — and the reason we quote a range after the audit, never a guaranteed rating.
03THE ORDER OF WORK
Appeals first,
volume second.
File the appeals
One evidence pack per review: what the customer claimed, why it breaks policy, dates, screenshots. Filed by us, decided by Google, tracked in one list you can read.
Rebuild from real customers
New reviews have to come from people who were actually served. We do not write them and we do not buy them: we fix what keeps customers from leaving one — a listing with wrong hours, no photos, an unanswered complaint — and the volume follows from there.
Report weekly
Rating, count, appeals won, replies sent. When the line is flat for a month, the honest move is to stop — not to sell you another quarter.
04QUESTIONS
Clear answers.
Before you pay.
Do replies change the rating?
No. A reply does not move the average by a single point. It changes what the next customer reads before deciding — which is why we still answer every review, including the ones we cannot appeal.
Can you delete a bad review?
Only if it breaks Google policy: spam, a conflict of interest, an off-topic rant, or a review from someone who was never a customer. A one-star review from a real customer who had a bad day stays. The fix is the average, not the review.
How fast can a rating move?
Arithmetic bounds it. On a profile with 120 reviews at 3.5, thirty published five-star reviews take it to 3.8. Nothing takes it to 4.6 without 330 of them, or a large number of successful appeals.
Is a fast flow of new reviews suspicious?
Not when the profile matches a local session: Android device, mobile carrier IP, coordinates on the point, human intervals. What gets filtered is a batch from one IP, not a business that finally started posting.
What if the profile only has a few reviews?
Then it recovers quickly: at twelve reviews, thirty-three published five-star reviews take a 3.5 rating to 4.6. Small profiles are fixed with requests, large ones with appeals and time.
05START WITH ONE POINT
Bring the number.
We do the math.
Send the Maps link. You get the current count, the entries that break policy, the rating without them and how many published reviews your target actually needs — before committing to anything.
Start with the
$50 pilot.
Five actions on one location: rating arithmetic, appeal list, field-check brief, reply set and a weekly report template. You approve every step before it runs.
Message the operator $50 PILOT · 5 ACTIONS