Reviews
How to Analyse Google Reviews and Turn Them Into Fixes
Learn how to analyse Google reviews with a simple spreadsheet: tag themes, spot recurring complaints, compare branches and competitors, and fix what matters.
RankMonk Team · · 10 min read
Your Google reviews already tell you what to fix: the slow billing counter, the receptionist who never picks up, the parking nobody can find. Most owners read reviews one at a time, reply, and move on, so the pattern never shows up. This guide shows you how to analyse Google reviews in an afternoon with a spreadsheet, find the two or three problems that cost you the most, and check whether your fixes worked.
You do not need special software for a single outlet. You need a list of themes, a few hours, and the discipline to act on what you find.
Why the star rating alone tells you very little
A 4.3 rating looks fine, but it hides the detail. Two clinics can both sit at 4.3: one gets praised for its doctors and criticised for waiting time, the other gets praised for speed and criticised for rude staff. They need completely different fixes.
The written text is where the reasons live. Google's own help page on tips to get more reviews says negative reviews can help you understand customer expectations and improve future experiences. Analysis is simply doing that systematically instead of by memory.
Reviews also matter for visibility. Google's article on how local results are ranked says local results are mainly based on relevance, distance and popularity, and that more reviews and positive ratings can help your local ranking. Fixing the problems behind bad reviews is one of the few ranking levers that also improves the business itself.
How to analyse Google reviews: a five-step method
This works for a salon with 80 reviews or a hospital with 3,000. Bigger businesses just spend more time on step one.
Step 1: Collect the reviews in one place
For one location, open your Business Profile, go to Read reviews, and copy each review into a spreadsheet with these columns:
- Date
- Star rating
- Review text
- Theme (you fill this in later)
- Positive or negative for that theme
- Replied? (yes or no)
Start with the last 6 to 12 months. Older reviews often describe a business you no longer run: a previous manager, an old menu, a location you moved from.
If you manage many locations, copying by hand stops being practical. Google's Business Profile API lets developers fetch review data in bulk, including the star rating, comment, create time and your reply, and it can pull reviews for several locations in one request. You (or your developer) need to register an application and get OAuth credentials first. Review management tools such as AI review replies and sentiment do this collection for you.
Step 2: Build a short theme list
Read 20 to 30 reviews before you tag anything, and write down what people keep talking about. Keep the final list to 8 to 12 themes, or the analysis turns into noise. A typical list looks like this:
| Business type | Example themes |
|---|---|
| Dental clinic | Waiting time, doctor's explanation, pain during treatment, pricing clarity, cleanliness, reception and calls, parking |
| Restaurant | Taste, portion size, delivery time, staff behaviour, cleanliness, value for money, ambience, billing errors |
| Salon | Result of the service, stylist skill, appointment delays, hygiene, upselling pressure, price |
| Gym | Crowding at peak hours, trainer quality, equipment condition, cleanliness, membership and refund issues |
| Coaching centre | Faculty, batch size, doubt-clearing, study material, fee and refund policy, communication with parents |
Use the words your customers use. If people in Andheri keep writing "waiting", call the theme "waiting", not "throughput".
Step 3: Tag every review
Go through each review and tag every theme it mentions, with a plus or minus. One review can carry several tags. "Doctor was excellent but we waited 90 minutes and nobody told us why" gets Doctor's explanation (+), Waiting time (−) and Reception and calls (−).
Two rules keep this honest:
- Tag what the review says, not what you think the customer meant.
- Tag rating-only reviews as "no text" and leave them out of the theme counts. They still count towards your rating, but they cannot tell you why.
Step 4: Count, then look at recency
Now count. For each theme, note how many positive and negative mentions it has, and how many of those came in the last 90 days. A simple pivot table does the job.
You are looking for three things:
- Your strengths. Themes that are overwhelmingly positive. Protect these and mention them in your profile description and posts.
- Your recurring problems. Themes with many negative mentions across months. These are structural, not bad luck.
- New problems. Themes that were quiet before and are now rising. A sudden cluster of "cold food" complaints usually means something changed: a new delivery partner, a new cook, a kitchen moved.
Weight recent reviews more. A complaint about parking from 2024 matters less if you added valet parking last year.
Step 5: Pick one or two fixes and give each an owner
Do not try to fix ten things. Pick the one or two negative themes that appear most often in recent reviews and that you can actually change. For each, write down:
- What exactly is going wrong (from the review text)
- Who owns the fix
- What will change, and by when
- What you expect to see in reviews if it works
Then put a date in your calendar 60 to 90 days out to repeat the count for that theme.
A worked example: a biryani outlet in Hyderabad
Say a biryani outlet in Gachibowli has 220 reviews and a 4.1 rating. The owner tags the last nine months, about 140 reviews, and finds this (illustrative numbers):
| Theme | Positive | Negative | Negative in last 90 days |
|---|---|---|---|
| Taste | 61 | 9 | 2 |
| Portion size | 12 | 18 | 4 |
| Delivery time | 3 | 21 | 13 |
| Staff behaviour | 19 | 6 | 1 |
| Billing errors | 0 | 7 | 5 |
Taste is clearly the strength. Portion size is a long-running grumble but steady. Delivery time is the real problem and it is getting worse, with most of the complaints in the last three months. Billing errors are small but new.
Reading the delivery reviews closely, most mention orders placed between 8 and 10 pm on weekends. That points to a kitchen capacity issue at peak hours, not a delivery partner problem. The fix: prep more handi in advance on Friday to Sunday evenings and pause delivery orders briefly when the dine-in queue is long. The billing complaints all mention a new POS system, so the manager retrains the cashiers that week.
Ninety days later the owner recounts only those two themes. That second count is what tells them whether the fix worked.
Read your competitors' reviews as well
Your competitors' reviews are public, and they show what customers in your area value and what nobody is doing well. If every gym in HSR Layout gets complaints about crowding between 6 and 8 pm, a gym that publishes its quiet hours or adds a batch-booking slot has a real reason to be chosen.
Take your three nearest competitors, read their last 50 to 100 reviews each, and use the same theme list. Look for:
- Complaints that appear at all of them (an unmet need in your area)
- Praise they get that you never do (a gap in your service or in how you describe it)
- Services customers mention that you offer but do not list on your profile
If you want to see which competitors outrank you for each search, a competitor tracker shows that alongside their review counts.
One firm rule: read competitors' reviews, never write them. Google's Maps user-contributed content policy prohibits reviews based on a conflict of interest, which includes industry competitors, and prohibits posting content on a competitor's place to undermine its reputation.
Comparing branches if you have more than one location
For chains and franchises, the same theme list run across every branch is one of the most useful reports you can build. It turns "Koramangala has a lower rating" into "Koramangala gets three times as many waiting-time complaints as Whitefield, on the same menu and the same prices."
A few things to keep in mind:
- Use one theme list for all branches, so the numbers compare like with like.
- Compare rates, not raw counts. A branch with 900 reviews will have more complaints than one with 90. Use negative mentions per 100 reviews.
- Separate brand issues from branch issues. If "price" is negative everywhere, that is a head-office decision. If "cleanliness" is negative at two branches, that is a branch manager's job.
- Share the results with branch managers, with the review quotes, not just the scores. Managers act faster on their customers' actual words.
Multi-location management tools help you read reviews across branches in one place instead of opening each profile.
Close the loop: reply, fix, then tell people
Analysis is only useful if customers see the result. Replying is part of that. According to Google's help page on managing customer reviews, the reviewer is notified when you reply and may update their review after reading it. When they update it, the review date changes to the date of the update.
So once you have fixed something, it is fair to say so in your replies to new reviews on that topic: "We have added two more billing counters since August, thank you for flagging this." Keep replies short, specific and honest, and only claim fixes you have actually made. If you need help with the wording, see our guide on how to respond to negative reviews.
Mistakes to avoid
- Turning analysis into review targets for staff. Google's content policy says merchants should not ask staff to solicit a certain number of reviews, and should not require or pressure customers to leave reviews while on the premises. Measure themes, not how many reviews each employee collected.
- Asking only happy customers. The same policy prohibits selectively soliciting positive reviews and discouraging negative ones. Selective asking also ruins your analysis, because the complaints you most need to see never arrive.
- Offering anything in return for a review, or for changing or removing one. Google's help page calls incentives in exchange for reviews strictly prohibited.
- Trusting an AI summary without reading the reviews. Summaries are a good starting point, but check every conclusion against the original text before you change how your business runs.
- Reacting to a single review. One furious review about a staff member deserves a reply and a conversation, not a policy change. Patterns deserve policy changes.
Frequently asked questions
How often should I analyse my Google reviews?
For most single-location businesses, a full analysis every quarter is enough, with a quick weekly read of new reviews in between. Businesses that get dozens of reviews a week, such as busy restaurants or hospitals, benefit from a monthly count.
How many reviews do I need before analysis is useful?
There is no official number. In practice, once you have 30 to 40 reviews with written text, repeated themes start to show. Below that, read every review carefully and treat any repeated complaint as worth checking.
Can I see the reviews of a competitor in my spreadsheet?
You can read and note them manually from their public Business Profile on Google Maps. The Business Profile API review data covers locations you manage, not other businesses.
Will fixing complaints improve my ranking on Google Maps?
Google does not promise that. What it does say is that more reviews and positive ratings can help your local ranking. Fixing what customers complain about is the honest way to earn better reviews over time.
Start with one afternoon
Export your last six months of reviews, build a list of about ten themes, tag every review, and count. Pick the one or two problems that show up most in recent months, give each an owner, and recount in 90 days. Repeat that every quarter and your reviews stop being a scoreboard you worry about and become a free, ongoing customer survey. If you would like to see how your reviews and rankings compare with competitors in your area, book a walkthrough with RankMonk.
- Google reviews
- Review analysis
- Customer feedback
- Operations
See how your business shows up
RankMonk tracks your rank street by street, protects your Google Business Profile and shows whether AI assistants recommend you.
Book a demo →