A bad review almost never comes out of nowhere. Providing good service is the whole job, and most of the time you deliver it. But things go wrong sometimes, and when they do, most customers do not run straight to Google. They call first.
They tell your team the job was wrong, that it was the second time, that they were "really not happy." That call is the moment the review can still be stopped. Then a few hours pass, nobody calls back, and they do the only thing left to do. They type it into Google or Yelp.
If you run a home-services business, your reputation is built or broken on the phone, not on the review site. The review is just the receipt.
And here's the brutal math: as an owner, there is no way I can listen to my calls. I am working on the business, not in it, the way most owners have to operate. I run a cleaning company with offshore agents taking fifty-plus calls a day, and I cannot sit and listen to fifty of them, much less even one. Like most owners, I am relying on my agents and employees to handle each call right.
The fact is, I want to be alerted when something goes wrong. Sometimes the agent is too busy handling the situation to flag it, and sometimes they would rather I never find out. That is exactly what Status Line is good at catching: the call where a good customer quietly decided they were done with us.
So the angry call sits in the log as a number and a duration. Unheard. And the window to fix it, the few hours between "I'm really frustrated" and a public one-star, closes while you're out on a job.
The review wasn't written on Google. It was written on the phone. Google was just where it got published.
The window you're missing
Reputation people will tell you the move is to respond fast to bad reviews. That's damage control. You're already bleeding. The real move is to never let the review get written in the first place. And there's a clean window to do exactly that:
Catch the call, call back inside that window, fix the problem, and the review never happens. The customer goes from "I'm telling everyone" to "they actually called me back and made it right," which is the story that turns an angry customer into a loyal one.
The only hard part is the catching. That's the part AI is now genuinely good at.
The fix: put AI on the line that already exists
You don't need to listen to every call. You need something that does, and only taps you on the shoulder for the ones that matter.
That's what we built Status Line to do. It connects to your Quo (formerly OpenPhone) business line, reads the full transcript of every call the moment it ends, and checks each one against rules you write in plain English. When a call looks like a bad-review risk, it emails you within minutes, with the exact quote that set it off and a button to call the customer straight back.
Heads up, a call just matched your alert.
"I'm afraid I need to cancel your service on the 22nd. I'm really sorry… you can find another cleaner or another cleaning company who can accommodate your request."
It doesn't match keywords. Here, customer support was about to cancel this cleaning because we were fully booked that day. It reads the whole call and understands the intent, then hands you the receipt.
No keywords to maintain. No scripts to tune. It understands what people mean. A caller who says "I'll be telling everyone how this went" gets flagged even though they never said the word "review."
Early access · 5-minute setup · Quo (formerly OpenPhone) customers
Steal these: 6 alert prompts that catch bad reviews
This is the part you actually want. An alert is just a sentence you write the way you'd explain it to a teammate. Below are six we've seen work for home-services owners, the obvious threats and the quiet ones that do the most damage. Copy any of them straight into Status Line and tweak the wording to sound like you.
A caller threatens a bad review
Alert me when a caller is angry or threatens to leave a bad review, mentions one star, Google, Yelp, Nextdoor, or says they'll tell people about a bad experience.
Get access to this alertAn unhappy customer who never complains
Alert me when a customer sounds disappointed, let down, or resigned about the service, even if they stay polite and don't complain directly. "It's fine, don't worry about it" said flatly counts.
Get access to this alertThe same problem, twice
Alert me when a caller says this is the second time, that it's happened before, or that they've already called about this and it wasn't fixed.
Get access to this alertThey ask for the owner
Alert me when a caller asks to speak to the owner or a manager, or says they want to escalate or take this further.
Get access to this alertA recurring client is slipping away
Alert me when a recurring client sounds like they're thinking of cancelling, pausing service, or switching to another company.
Get access to this alertA refund or make-good request
Alert me when a caller asks for a refund, asks not to be charged, or wants a discount to make up for a problem with the job.
Get access to this alertThe quiet ones matter most. The customer who yells will sometimes call back and give you a chance. The one who says "no, it's fine, don't worry about it" in a flat, done voice. That is the customer who hangs up and writes the review without warning. A human skimming a transcript misses that. AI reading every word catches the resignation.
Three calls it would catch
These are the kinds of calls Status Line flags, drawn from real patterns on a working cleaning company's Quo line. Not one of them says the words "bad review." All three are exactly how one starts.
"I'm afraid I need to cancel your service on the 22nd. I'm really sorry… you can find another cleaner or another cleaning company who can accommodate your request."
Your own customer support team tried to cancel a booked customer's service and pointed her at a competitor, in a 40-second call. Classified: existing customer, sale lost, negative tone.
That's a one-star review writing itself, and the kind of call an owner never hears. It was flagged the same day so I could address it with the customer support team and work with them to get this customer booked on another day.
"No, it's okay… don't worry about it. It's fine. I'll just sort something out myself."
No raised voice, no threat, no demand. Just a customer who went flat and gave up on us mid-call. Classified: existing customer, dissatisfied, polite-but-negative tone.
The quiet ones are the dangerous ones. She won't argue. She'll just stop booking and leave a calm, lethal review later. A human skims a short, polite call and moves on. Flagged, you get the one chance to call back and find out what really went wrong.
"The apartment was supposed to be move-out ready, and now my landlord is refusing to give back my deposit because of how it was left. I am really not happy with how this was cleaned."
A customer furious about the cleaning quality on a move-out job, with her landlord threatening to withhold her deposit because of it. Classified: existing customer, service failure, highly negative tone.
This is the severe one. A botched job with real money on the line is how a five-star regular becomes a public one-star and a refund demand in the same week. Flagged the moment the call ended, you can step in, make it right, and protect both the customer and her deposit before any of it reaches Google.
How to set it up in five minutes
Connect your Quo line
Paste your Quo (formerly OpenPhone) API key. Status Line starts reading and classifying every call the moment it ends, and backfills your recent history so you have data to test against straight away.
Write a rule in plain English
Open Alerts, hit new rule, and describe what to watch for. Paste one of the six above. No keywords, no logic builder. You're just telling it what matters to you.
Backtest before it goes live
Run the rule against your last 50 calls and watch what it would have flagged, with the quote and a confidence score for each. Tune the wording until it's catching the right calls, then flip it on. No guessing, no crying wolf.
Get the alert. Make the callback.
From then on, any matching call emails you within minutes and lands in your dashboard with a one-tap call-back button. You do the one thing that saves the review: pick up the phone while it's still warm.
You can't listen to every call. Now you don't have to. You just have to answer the one that matters, before it becomes a one-star.
Alerts are just the start
Catching bad-review risk is one thing Status Line does, not the whole of it. It's call intelligence for your entire operation, turning every conversation on your line into something you can actually use. You can ask the AI about your calls in plain English:
- What are our top five objections?
- Why did we lose business this week?
- Who are my top-performing agents, and why?
And it all rolls up into a metrics dashboard, so the patterns across hundreds of calls are sitting right in front of you without you listening to a single one. The bad-review alert is the first thing owners feel, but it's one feature of many.