Most owners treat reviews like a report card: five stars is good, one star stings. But reviews on Google Maps and other sites are also free data. Customers tell you, in their own words, where your process broke down. You just have to read them all together instead of one at a time, and look for what repeats.
I do this kind of review analysis for small business owners, and I see the same thing almost every time: different complaints usually come down to the same two or three causes. And most often they're not about the quality of the service itself, but about everything around it.
Why reviews are about money
Every recurring complaint stands for customers who probably didn't come back and people who read the review and never showed up. Praise, on the other hand, tells you why people choose you and what you need to protect when you change things. A review won't tell you exactly how much you're losing, but it will tell you where to look.
Step 1. Gather a year of reviews
Collect the reviews from the last 12 months on every site where you have them. If there aren't many, go back two years. The most useful ones aren't just the one-star reviews but the three- and four-star ones, too: that's where customers calmly explain what went wrong even though they were mostly happy.
Step 2. Sort complaints by process, not by emotion
Don't argue with the review or look for someone to blame. For each complaint, ask one question: at what stage of the work did this happen? For a service business, almost every complaint falls into a few groups:
- Contact and booking: slow replies, couldn't get through, wrong appointment time.
- Waiting: showed up on time and waited half an hour.
- Price surprises: the final bill was higher than quoted.
- Communication: nothing was explained, staff weren't attentive, different people said different things.
- Service result: not what they expected.
- After the visit: nobody called back, the problem wasn't solved, the complaint was ignored.
Step 3. Count the repeats
Next to each group, write a number: how many times it shows up. One complaint is a one-off. Three or four similar ones in a year point to a recurring problem in the process. Separately, note what people complained about in the last three months. That tells you what hurts now, not a year ago.
Step 4. Find the cause in the process
For each of the two or three most common groups, ask: what process is creating this? For example:
- "Waited forever in the lobby" often means the schedule doesn't account for how long services actually take, or for late arrivals.
- "Got charged more" often means the full price isn't given at booking, or there's no rule for warning customers about extras.
- "Nobody called me back" often means incoming inquiries have no owner and no status.
These are hypotheses, not a diagnosis. You can check them against your own data: the schedule, the messages, the bookings.
Step 5. Decide what to fix first
Pick just one cause: the one that's come up most often in recent months and is within your control, not driven by outside factors. Write a rule that fixes it, assign a person to own it, and in two or three months check whether that complaint shows up less often.
Don't skip the praise
If customers often praise a particular stylist, a thoughtful receptionist, or the fact that "they explained everything," those are your strengths. Protect them when you change processes, and use them in your marketing and on your website.
Read your competitors' reviews too
It's worth running the same analysis on two or three similar businesses near you. If their customers keep complaining about slow replies or surprise charges, and you have that under control, that's an advantage worth highlighting. And if they're praised for something you don't offer, you've learned what customers in your area expect.
Don't stop at public reviews, either. Complaints in your messages, reasons for cancellations, and the questions customers ask over and over are the same kind of data. You're just the only one who can see it.
Reply to reviews
Your reply to a complaint isn't read only by the person who wrote it. Every future customer reads it too. Thank them briefly, acknowledge the problem without making excuses, and say what you changed. And never buy fake reviews: in the EU they're explicitly banned under Directive (EU) 2019/2161, and customers usually spot them anyway.
Review analysis checklist
- You've gathered reviews from the last 12 months on every site.
- You've read not just the bad ones but the three- and four-star reviews too.
- Each complaint is assigned to a stage of the work, not to a person.
- You've counted how often each group appears, with the last three months counted separately.
- For the top two or three groups, you've written down a hypothesis about the process behind them.
- You've chosen one cause and a rule that fixes it.
- You've noted the strengths you need to protect.
Or let me do it for you
If you'd rather not spend an evening on this, the LACIFRA website offers a free sample on your own business: a review analysis. Send me your Google Maps link, or your business name and city. I'll go through your public reviews and send you the three most common problems, the process behind each one, and what to fix first. The report arrives by email.