Read your Google reviews to find the real problems, and only fight the fake ones
Not all negative reviews are equal. A simple method to sort, in twenty minutes, what is a problem to fix, a reply to write, or a report to file.
When the rating drops, the reflex is to look at the stars. It is the worst indicator. A profile at 4.3 can hide a real service problem, a coordinated attack, or simply three grumpy customers in a quiet month. The three situations call for different responses, and confusing them costs time and money.
Here is the method we apply when a business owner sends us their profile.
Step 1: export before you interpret
Open the Business Profile, reviews section, and list the last twelve months of reviews in a table: date, score, author, text, replied or not. A spreadsheet is enough. You will immediately see things the Maps page hides: series, gaps, recurring authors.
If you have more than 300 reviews, stick to the last six months.
Step 2: sort the negatives into three columns
For each one-, two- or three-star review, ask one question: was this person a customer?
Column A: real customer, specific complaint
“40-minute wait on Saturday”, “quote exceeded by 300 euros”, “never called back”. These reviews are free information about how you operate. They cannot be removed and should not be. They call for a reply and, often, a fix.
Column B: real customer, vague or disproportionate complaint
“Rubbish”, “disappointed”, “not worth the price”. A real customer, an opinion. Not removable. A short courteous reply is enough, and your flow of positive reviews will do the rest.
Column C: no trace of the customer, or warning signs
No customer by that name, a described service you do not offer, a recently created account, a series of reviews on the same day, trade vocabulary, a recommendation for a competitor, remarks about a named person, a demand for money. These are the reviews to analyse for a report.
Step 3: look for patterns in column A
Group the specific complaints by theme: delays, price, welcome, quality, communication. If one theme accounts for more than a third of negative reviews, you have an identified operational problem, and that is the best possible news: it can be fixed, and the fix shows up in the next reviews.
Look at days and times too. Negative reviews clustered at weekends or in the evening usually point to a staffing gap rather than a deep problem.
Step 4: read the positives as well
Five-star reviews tell you what your customers value, in their words. Those are the words to reuse in your communication, and the ones prospects search for. If “fast” comes up twenty times, that is your argument, not “premium quality”.
Step 5: check the rhythm
Count reviews per month. A healthy business receives a steady flow. Two warning signs:
- A spike of negatives over a few days: probable coordinated attack, a case to build fast.
- A long gap: you have stopped asking, and your rating becomes fragile against the next unhappy customer.
What a proper sort produces
On a typical profile at 4.3 with 80 reviews, we generally find:
| Category | Share of negatives | Action |
|---|---|---|
| Real customers, specific complaint | 40 to 60% | Reply, fix |
| Real customers, vague complaint | 20 to 30% | Reply briefly |
| Suspicious or against the rules | 15 to 30% | Report with a file |
In other words, about a quarter of the negative reviews on an average profile are contestable, and the remaining three quarters are a source of improvement. Working both fronts at once is what lifts a rating durably.
The trap to avoid
Trying to get column A or B reviews removed. They will not go, the appeal will be wasted, and you will have lost the information they contained. The energy goes into column C, and only there.
If you want a second opinion on your sort, send us the profile URL. We identify for free the reviews that have a real chance of being taken down.