When a pipe bursts at 2pm on a Tuesday, the homeowner calls the first plumber they find. If no one picks up, they call the second.
The first plumber is on a job. Hands dirty. Phone in pocket, ringing.
They lose the customer not because their work is bad. Not because their price is wrong. They lose it because they were busy doing the actual job they built their business to do.
This is the front desk problem — and it costs service businesses $35,000+ a year.
The Math Nobody Talks About
Here's a rough calculation I keep coming back to:
- Average missed calls per day for a busy service business: 4-6
- Conversion rate if you answer: ~40%
- Average customer lifetime value (plumbing, HVAC, salon): $1,200–$3,500
Miss 3 calls a day, 5 days a week. Convert even half. You're looking at $2,500–$5,000/month in revenue that evaporates because the phone rang at the wrong time.
Hiring a receptionist is the obvious fix. At $35K/year fully loaded, it's also an obvious problem for a small operator running 2–3 crews.
Answering services exist. They charge $245+/month, read from a script, can't actually book anything, and have a nasty habit of getting your customers' names wrong.
Neither scales. Neither integrates with how small operators actually run.
What AI Should Actually Be Doing
There's a lot of AI hype right now. Most of it is solving problems that didn't need solving.
But when I look at a plumber who misses 5 calls a day while under a sink — that's a real workflow problem. The kind AI can actually fix.
The technical stack isn't magic. It's a few components working together:
1. Natural language understanding
A caller says "my AC is making a grinding noise and it's 95 degrees outside, I need someone today." The system extracts: emergency HVAC, same-day request, likely urgent priority.
2. Intent classification
Is this a booking? A quote request? A callback? An existing customer with a question? Different intents route differently.
3. Calendar integration with conflict detection
Actually check availability. Don't just say "we'll call you back." Book the appointment, check for conflicts, confirm the slot. The customer leaves the conversation with a confirmed time.
4. Automated follow-ups
Day-of reminder. Post-job follow-up. Review request. All triggered automatically without the owner touching anything.
This is the workflow. It's not complicated to describe. The hard part is building it well enough that real business owners trust it with their front door.
What We Built
FrontHawk is an AI receptionist for service businesses — plumbers, HVAC technicians, electricians, salons, contractors.
It handles the full front-desk workflow:
- Answers inquiries 24/7 via phone and web, in natural language
- Qualifies leads — captures name, issue, urgency, preferred time
- Books appointments directly into your calendar with conflict detection
- Sends confirmations and reminders automatically
- Requests reviews after completed jobs
The owner wakes up to a booked schedule instead of a voicemail inbox.
A few things we got specific about during the build:
Tone calibration by industry. A salon conversation sounds different from an emergency plumbing call. The urgency level, the language, what information you capture first — all different. Generic doesn't work here.
Graceful handoff. When something is genuinely outside what the AI should handle (complex diagnosis, unhappy customer, anything requiring judgment), it escalates cleanly rather than hallucinating an answer.
No new software for the owner. Calendar integration syncs to what they already use. The owner doesn't learn a new system — they just have fewer missed calls.
The Part That Surprised Us
When we started testing with real businesses, the biggest win wasn't the missed calls.
It was the after-hours bookings.
Homeowners search for plumbers at 9pm when they notice a leak. They search for HVAC service at 7am when the AC didn't turn on overnight. They want to book a hair appointment on Sunday while they're thinking about it.
The business is closed. No one answers. They move on.
FrontHawk is running at 2am. The booking lands on the owner's calendar. They wake up to confirmed work they never would have seen otherwise.
Where It Fits in the Stack
For the technical readers: FrontHawk integrates with standard calendar APIs (Google Calendar, Acuity, Calendly-style systems), handles webhook-based call routing, and uses structured extraction to pull appointment data into whatever CRM or scheduling tool the business already uses.
The AI layer does the conversation. The integration layer does the actual work. The owner sees results, not infrastructure.
Pricing
- Starter: $49/month — 7-day free trial, covers most small operators
- Pro: $79/month — higher volume, multi-location support, advanced analytics
No setup fees. No per-call charges. Flat rate.
AI that generates marketing copy for businesses that don't need it gets a lot of press. AI that catches a phone call while a plumber is under a sink — that's the use case I care about.
If you run a service business (or know someone who does), try it free for 7 days → https://fronthawk.polsia.app
Top comments (0)