Plumbing Example: Booking 22 Jobs Without Taking After-Hours Calls

This is an illustrative scenario, not a client result. The business, people and numbers are examples chosen to show how the math works. Replace them with your own.

Imagine an owner who started a plumbing company 11 years ago. Two crews, a solid Google rating, and more work than he can comfortably handle in peak months. From the outside, a successful small business.

From the inside: the owner is the phone. Every call in, every call out, every booking, every question about availability. While he is on a job site — which is where he spends most of his day — calls pile up. After 6pm, they go to voicemail and stay there.

He finishes a job, sees missed calls, and returns some. Several don’t pick up, and by morning some have hired someone else.

In this example, we assume that in the 30 days after setting up AI intake, 22 jobs are booked from leads the owner never personally spoke to during intake.

The Problem Was the Owner, Not the Business

This is a common pattern with owner-operated service businesses: the owner becomes the bottleneck.

The company isn’t losing jobs because of bad service or poor reputation. It loses them because the owner physically can’t answer the phone with both hands on a job — and because a homeowner with a slow drain or a dripping pipe often won’t wait three hours for a callback. They search “plumber near me” and call the next listing.

Before Numbers (Assumptions)

For this example, assume 60 days of call data from the phone provider show a typical month like this:

24 leads a month, gone. If every one had become a job at $340, that is up to $8,160 a month. Not every lead converts, so treat that as a ceiling.

The emergency calls hurt the most. Burst pipes, water heater failures, flooding — those callers rarely leave a voicemail. They call the next plumber. For a company whose website is part of the problem, see this plumbing lead generation worked example.

What the Build Could Look Like

An intake audit would run first: mapping where calls come from, what questions the owner asks to qualify a job, and how he dispatches across two crews.

Three things would shape a build like this:

1. Emergency escalation is non-negotiable. Any call involving flooding, no hot water in a home with young children, or potential pipe failure needs to reach the owner immediately — not queue until morning. That means a separate emergency path with a real-time text alert to his cell. See also the 11pm burst pipe scenario.

2. Service area filtering. Most crews have zip codes they cover well and areas they don’t. The AI asks for the service address early and confirms it’s within range before investing further in intake.

3. Booking into real crew availability. If the owner uses a simple calendar to manage two crews, the AI can connect to it and offer actual available windows — not a generic “we’ll call you to schedule.” The customer picks a slot. It’s confirmed. No phone tag.

For this example, assume about two and a half weeks from kickoff to live.

Month One: What the Numbers Could Look Like

Assume the system goes live on a Monday with no announcement to customers.

Assumed 30-day numbers:

Revenue from AI-captured bookings:

• Standard jobs: 16 × $340 average = $5,440
• Emergency jobs: 6 × $555 average = $3,330
• Total: $8,770 in booked revenue from leads the owner didn’t personally field

Not all of that is new money. Before AI intake, callbacks recovered about 8 leads a month. For a conservative view, count only 14 of the 22 jobs as new. At the blended average of about $399 a job ($8,770 ÷ 22), that is about $5,580 a month of incremental revenue.

Saturday-night emergencies matter most: a burst pipe answered in seconds can become a same-day job.

What Could Change Week Over Week

The first week can feel strange: calls that used to interrupt the day now show up as bookings in the calendar.

By week three, the operational shift would likely be more concrete:

The morning routine changes. Instead of starting the day returning missed calls from the night before, the owner opens the calendar to confirmed bookings and the CRM to a ranked list of leads that need a personal touch — complex jobs, large commercial inquiries, customers who specifically asked to speak with someone.

He stops being the dispatcher. The AI schedules both crews within his rules, and he reviews the calendar on Sunday evening.

The office assistant can shift focus. Call-return hours can move to customer check-ins and review requests.

For this example, assume the owner tracks his time for two weeks and finds 8 hours a week recovered from phone management. If he bills $115 an hour on direct work, that is 8 × $115 = $920 a week in time he could put back into paid work.

The Investment Conversation

Off-the-shelf AI receptionist products are inexpensive: as of October 2026, Goodcall lists plans from $79 to $299 a month and Smith.ai lists paid AI receptionist plans from $150 to $500 a month base. Managed systems with emergency routing and crew scheduling cost more.

For this example, assume a managed AI intake system costs $10,000 to set up and $1,500 a month.

How fast that pays back depends on which number you trust:

• All 22 jobs, on revenue: $8,770 − $1,500 = $7,270 a month toward setup. $10,000 ÷ $7,270 ≈ 1.4 months, or about six weeks.
• Only the 14 incremental jobs, on revenue: $5,580 − $1,500 = $4,080 a month. Payback ≈ 2.5 months.
• All 22 jobs, on gross profit at an assumed 50% margin: $4,385 − $1,500 = $2,885 a month. Payback ≈ 3.5 months.

The year-two cost of $18,000 is about two months of the example’s booked revenue. Your model should use your own call logs and margins.

Is Your Service Business in the Same Position?

If you’re an owner-operator handling your own phones, or running a small team where intake falls through the cracks after hours, this pattern is probably familiar. See how AI intake fits local service businesses.

The math question is always the same: what are you losing each month to unanswered calls, and what would it take to get that back? The HVAC version of this example uses the same math.

Our AI Readiness Audit answers that question for your business — two weeks and a written ROI model with your actual numbers. The 2-week AI Readiness Audit starts at $2,500, and the fee credits toward the setup of a build started within 90 days (up to 40% of the setup price).

See what the audit includes for local service businesses →