Case Study Reputation Management

Review Request &
Reputation Automation

Every completed job triggers an AI that reads customer sentiment and either sends a personalized Google review request — or quietly alerts the manager before a bad review goes public.

The Problem

Good work. Zero reviews. One bad one.

Most local service businesses do great work. Their customers are happy. But nobody asks for reviews — it feels awkward, so it never happens.

Meanwhile, the one customer who had a bad experience? They leave a review immediately. Without intervention, that 1-star sits on your Google profile for years, dragging down an otherwise stellar reputation.

The result: 34 total reviews, a 3.8-star average, and a steady stream of potential customers picking a competitor with more social proof — even though your work is better.

34
Google Reviews
Before automation
3.8★
Average Rating
Dragged down by 2 bad reviews
What was happening after every job:
  • Happy customers left. No follow-up.
  • Unhappy customers left. Left a 1-star review 3 days later.
  • Owner found out when a new customer mentioned it on the phone.
The Solution

Automated Sentiment Routing — Post Every Job

How it works — every single time

🔧
Job Complete
Tech closes ticket
Trigger
Make.com / Webhook
📱
Feedback Request
SMS via Twilio
🤖
GPT-4o
Sentiment analysis
Route A
Review Request
Personalized SMS + email sent to customer
⚠️
Route B
Manager Alert
Issue flagged privately. No review ask.
↗️
Route C
Recovery Message
Follow-up sent. Review ask deferred.
The Results

34 reviews to 91. 3.8 stars to 4.6.

Within 90 days of deploying the automation, a local HVAC company went from 34 Google reviews to 91 — without anyone on staff manually asking a single customer.

More importantly: two complaints that would have become public 1-star reviews were caught by the manager alert system. Both customers received a call within 24 hours. Neither posted a negative review.

The 4.6-star profile now ranks above three competitors in local search — generating an estimated 8–12 additional inbound calls per month.

91
Google Reviews
+57 in 90 days
4.6★
Average Rating
Up from 3.8
2
Bad Reviews Caught
Before going public
#1
Local Search Rank
Above 3 competitors

Under the Hood

🤖
GPT-4o — Sentiment + Routing
Reads stars + written feedback together. Returns structured JSON with routing decision, personalized message copy, and manager summary — in a single API call.
Make.com — Automation Backbone
Connects the job management software (ServiceTitan, Jobber, etc.) to the AI routing system. Triggers on job close, passes context, routes the output.
📱
Twilio — SMS Delivery
Sends the review request or recovery message as an SMS from a local number. Response rates are 3–5× higher than email alone.
🐍
Flask — API + Demo Backend
Lightweight Python backend handles the sentiment analysis API, serves the demo, and enforces rate limiting.

Works For

🔧
HVAC
After every service call
🚿
Plumbing
After repair or install
🦷
Dental
After each appointment
🍽️
Restaurant
After table turnover
🚗
Auto Service
After pickup or delivery
Any business that closes jobs can use this.

If you use Jobber, ServiceTitan, Jane App, OpenTable, or any job management tool with a webhook — this can be wired in within a day.

See It In Action

The demo below lets you simulate both sides: the customer submitting feedback, and the business generating a manual review request. The routing is live — real GPT-4o, real output.