A two-truck plumbing operation in Phoenix was drowning in scheduling calls, estimate follow-ups, and invoice chasing. Here is how AI automation changed the math without replacing a single person.
Mike Sandoval runs a two-truck plumbing business in Phoenix. He started the company nine years ago with one truck and a lot of referrals from neighbors. By 2024, he had two trucks, four employees, and a problem: the business was eating him alive.
Not the plumbing work. The paperwork.
"I was spending probably three hours every morning just on the phone and texting," he told us. "Scheduling, confirming appointments, answering the same questions about pricing. By the time I got to an actual job it was already 10 AM."
He was also losing work. Homeowners who called after hours or on weekends would try two or three other plumbers while Mike's phone went to voicemail. By the time he called back the next morning, they had already booked someone else.
In October 2024, Mike got a call from a property management company managing 47 units in Scottsdale. They needed a reliable plumber for routine calls across their portfolio. Mike wanted the contract badly. But when they asked how quickly he could respond to service requests, he had to be honest: during the day, within a few hours. Evenings and weekends, the next morning.
They went with someone else.
That was the moment Mike decided to actually look at what AI could do for a trade business his size.
Mike worked with an AI implementation partner over about six weeks. The total cost was $2,800 for setup and $350 per month to run. Here is what they built.
An AI-powered intake system that handles every inbound call and text message. When a homeowner reaches out, the AI responds immediately, asks the right qualifying questions (type of issue, urgency, property address, availability), and either books the appointment directly into Mike's scheduling software or routes it to him for a callback if the situation requires more judgment.
The system also handles appointment confirmations, sends reminder messages the day before and two hours before, and processes reschedule requests automatically. Mike still reviews the schedule every morning, but he is no longer the scheduling bottleneck.
They also set up automated follow-up for estimates that had not converted. When Mike sends an estimate that goes unanswered for 48 hours, the system sends a polite follow-up message. If it goes unanswered for five days, a second one. Mike was too busy to do this manually. The AI does it without him thinking about it.
Finally, they automated invoice reminders. Commercial clients who had not paid within 15 days got a professional reminder. Within 30 days, a second one. Collections improved. Cash flow improved.
Mike tracked his time carefully for the three months before and three months after implementation. The before: approximately 22 to 25 hours per week on administrative tasks. The after: 4 to 6 hours, mostly reviewing what the AI handled and making judgment calls the system flagged for him.
The 20-hour weekly reduction is not hypothetical. It is documented in his own time logs.
Beyond time: the property management company in Scottsdale called back. Mike pitched them again. This time, he could honestly say his system responded to every service request within minutes, around the clock. He got the contract. At roughly 15 calls per month at an average ticket of $280, that is $4,200 per month in new recurring revenue.
The AI system paid for itself in the first week of the Scottsdale contract.
Mike still answers his phone for complex situations. He still personally follows up with customers after big jobs to make sure they are happy. He still makes every judgment call about pricing and what work to take on.
The AI handles the volume and the routine. Mike handles the craft and the relationships. That division of labor is exactly how it is supposed to work.
"I thought it was going to be some complicated tech thing I had to manage. It is not. I looked at it for about 15 minutes after they set it up and then basically forgot about it. It just runs."
"The thing that surprised me most was how fast it responds. Someone texts at 11 PM on a Saturday and they get an answer in like 30 seconds. I used to lose those calls every single time."
For a two-truck plumbing operation in Phoenix, the ROI on AI automation was faster, more concrete, and more significant than Mike expected. The 20 hours he reclaimed per week are now split between taking on more work and getting home earlier.
Both outcomes, he said, were worth it.
*Some links above may be affiliate links. We only recommend tools we actually use.*
Tell us what is costing you the most time. We will map out exactly what your business needs. Free, no obligation.
Get Started Free