Case Study: How a Plumbing Company Cut Missed Calls by 85% with an AI Voice Agent

Note: This case study is an illustrative example based on composite data from real client engagements and industry benchmarks. Names and specific details have been changed. The results are representative of what is achievable with a properly configured AI voice agent in a service business.

The Problem

"Gulf Coast Plumbing" is a 6-person plumbing company based in the Houston metro area. Owner Mike runs a crew of four plumbers plus one office manager, Linda, who handles phones, scheduling, invoicing, and customer follow-ups.

Here is the problem Mike did not realize he had: he was missing 62% of inbound phone calls.

That is not unusual. According to industry data, service businesses miss an average of 49% of incoming calls (source). For small shops without a dedicated call center, it is even higher.

Think about that. For every 10 people who call a plumbing company, 5-6 get voicemail. And the vast majority of callers who reach voicemail don't leave a message -- they just call the next company on the list.

Mike's situation:

Linda handles phones from 8 AM to 5 PM, but she is also doing scheduling, invoicing, and email

When Linda is on a call, other calls go to voicemail

After 5 PM, weekends, and holidays -- 100% voicemail

Emergency plumbing calls at 9 PM on a Saturday? Straight to voicemail

Mike estimated he was losing $8,000-$12,000 per month in potential revenue from missed calls. That is conservative.

The Solution

We deployed an AI voice agent configured specifically for Gulf Coast Plumbing. Here is what it does:

Call Handling

Answers every call within 2 rings, 24 hours a day, 365 days a year
Greets callers naturally (not robotic -- think "friendly receptionist," not "press 1 for...")
Identifies the caller's issue through conversation
Determines urgency (emergency vs. routine)

Qualification and Booking

Asks qualifying questions: service area, type of issue, property type (residential/commercial)
Checks the real-time schedule (synced with their Google Calendar)
Books appointments directly into open slots
For emergencies, triggers an alert to the on-call plumber's phone

Follow-Up

Sends confirmation texts to customers immediately after booking
Sends appointment reminders 24 hours before the job
After the job, sends an automated review request with a Google Reviews link

Handoff

If the AI cannot handle a question (pricing for complex jobs, warranty disputes, etc.), it takes a message and flags it as "needs human follow-up"
Linda gets a morning summary email with all overnight calls, bookings, and flagged items

The Setup

Timeline: 2 weeks from kickoff to live deployment

Week 1:

Recorded 30+ real phone calls (with customer permission) to understand common scenarios

Built out the AI's knowledge base: services offered, pricing ranges, service area, scheduling rules, emergency protocols

Configured integrations: phone system, Google Calendar, CRM, SMS platform

Week 2:

Internal testing with the team making test calls

Refined responses based on edge cases

Soft launch with the AI handling after-hours calls only

Full deployment after 3 days of clean after-hours operation

Cost:

Setup: one-time configuration fee

Monthly: approximately $200/month for the voice AI platform, phone number, and SMS

Compare that to: hiring a second receptionist ($2,800-$3,500/month) or an answering service ($500-$1,500/month with limited capabilities)

The Results (First 90 Days)

| Metric | Before AI | After AI (90 days) | Change |
|--------|-----------|-------------------|--------|
| Calls answered | 38% | 97% | +155% |
| Missed calls | 62% | 3%* | -85% reduction |
| Appointments booked per week | 22 | 34 | +55% |
| After-hours bookings | 0 | 8/week avg | New revenue |
| Google reviews (monthly) | 3-4 | 14-16 | +300% |
| Estimated monthly revenue impact | -- | +$9,200 | -- |

*The 3% of "missed" calls were edge cases: prank calls, wrong numbers, and calls in languages the AI was not configured for (Spanish support was added in month 2).

Linda's Role Changed

Linda did not lose her job. She got a better one. Instead of being chained to the phone, she now focuses on:

Complex customer situations the AI flags

Invoicing and accounts receivable (which was falling behind)

Vendor management and supply ordering

Actually taking a lunch break

Mike told me: "Linda used to be stressed every day. Now she comes in, reviews the overnight summary, handles the stuff that needs a human, and actually has time to do her real job."

What Went Wrong

Not everything was smooth. Transparency matters, so here is what we had to fix:

The AI initially quoted prices too specifically. Plumbing jobs vary wildly based on what you find behind the wall. We had to retrain it to give ranges and say "we will provide an exact quote on-site."

One customer complained the AI was "too friendly." Seriously. The AI was cheerful on a call about a sewage backup. We adjusted the tone detection to match urgency.

Spanish-speaking callers were getting English responses. Houston has a massive Spanish-speaking population. We added Spanish language support in month 2 -- should have been there from day one. Lesson learned.

The AI booked a job outside the service area. It did not know that "Conroe" was a 90-minute drive. We added a zip code and city validation layer.

Every one of these was fixable. But they illustrate why AI agents need monitoring and iteration -- especially in the first 30 days. (For more on AI risks and how to mitigate them, read 5 Dangers of AI Agents.)

The Takeaway

Mike's plumbing company went from losing an estimated $8K-$12K/month in missed calls to capturing almost all of them -- for about $200/month.

The AI does not replace his team. It makes his team more effective. Linda is happier and more productive. His plumbers are busier. His customers get faster responses. His Google reviews are climbing, which feeds more inbound calls.

And Mike told me the best part: "I went on vacation for the first time in three years. The phones didn't stop. The bookings didn't stop. I actually relaxed."

That is what AI agents are supposed to do. Not replace humans -- remove the bottlenecks that prevent humans from doing their best work.

Want to see what an AI voice agent would look like for your business? Every service business is different -- I will map out a setup specific to your situation. Book a free consultation at aiguyjosh.com/contact.