Most small businesses have the same quiet frustration. Customers say lovely things on the day, promise to leave a review, and then life gets in the way. Months later the Google profile still shows a handful of reviews, and one grumpy one sits near the top.
Nobody is doing anything wrong. Asking for reviews is just one more job that depends on someone remembering, at the right moment, to send the right message. That is exactly the kind of job a small, well-built AI system handles well.
Why review requests fall through the cracks
In our experience, the problem is rarely that customers are unhappy. It is usually that:
- The job finishes and everyone moves straight on to the next one
- The request goes out weeks later, when the good feeling has faded
- The message is a generic template that reads like it came from a robot
- The link is buried, so leaving a review takes five taps instead of one
What a sensible AI setup does
A good review request system connects to the tools you already use, such as your job management software, booking system or CRM, and does three things.
It notices when a job is finished. When a job is marked complete, a settlement goes through or an appointment wraps up, the system queues a request. Nobody has to remember.
It writes a personal message. Using details from the job, it drafts a short, warm note in your voice. "Thanks for having us out to fix the hot water system on Tuesday" lands very differently to "Dear valued customer". Each message includes a direct link to your Google review page.
It sends one gentle reminder, then stops. If there is no response after a few days, one friendly follow-up goes out. After that, the customer is left alone.
Staying fair and within the rules
This part matters, and it is where DIY setups often go wrong.
- Ask everyone, not just the happy ones. Google's review policies do not allow businesses to selectively ask only customers they expect to be positive, and the ACCC has published guidance on online reviews for businesses that covers the same ground. Filtering out unhappy customers before asking is called review gating, and we do not build it.
- No incentives. No discounts, prizes or freebies in exchange for a review.
- No fake or AI-written reviews. The system writes the request, never the review. Customers write their own words.
- Respect opt-outs. If someone asks not to be contacted, they are taken off the list straight away.
Where a human stays in charge
We always build in review points. You decide which jobs trigger a request, and anything unusual, like a job with a complaint or a refund in progress, is held back for a person to look at first.
When a negative review does come in, the system can flag it and draft a calm, professional reply for you to edit. It never posts a response on its own, and it never argues with a customer in public.
Customer contact details stay in your systems. Our plain-English privacy checklist covers the questions we ask before connecting anything.
What it will not fix
Honest limits matter here too. A review system will not turn a poor experience into a five-star review, and it should not try. Fix the service first.
It also only works if jobs are marked complete in your software. If finished jobs live in someone's head rather than the system, requests will not go out.
Is this the right place to start?
For many trade and service businesses, review requests sit alongside chasing overdue invoices and catching missed calls as one of the simplest wins. The rules are clear, the work is repetitive, and the result shows up where new customers look first. It suits a Quick Win Build when the process is already clear, or a line item in an AI Opportunity Audit if you want to weigh it against other ideas.
At Cloud Inn we design, build and run practical AI systems for local businesses without an IT department, on fixed fees and with a human in the loop. If your reviews do not reflect the work you do, book a 30-minute call and we will tell you honestly whether it is worth automating.