Built for your business
AI Chat Assistant for Plumbers, Electricians & HVAC — Triage 24/7
Catch the 11pm burst-pipe call, triage the real emergencies from the wait-til-mornings, and answer the same five FAQ questions all day — without your dispatcher fielding any of it.
The problem
A home services business runs on the phone. When the phone is answered fast and the right calls get to the right techs, the business prints money. When the phone is buried — by FAQ questions your dispatcher has answered three hundred times already, by voicemails left between 9pm and 6am, by callers who get tired of waiting and dial the next contractor on the search results — the trucks run light and nobody can quite explain why the week felt slow.
The math behind why this hurts is brutal. Housecall Pro's analysis of inbound call patterns found that home service businesses miss roughly 27% of their inbound calls on average, and pegs the average missed call at about $1,200 in lost revenue. Invoca's 2025 Home Services Call Conversion Benchmarks Report, built from over 60 million phone calls, found that only 55% of callers actually reach a live person, 37% of calls from digital marketing qualify as leads, and 46% of those qualified leads convert during the call itself. In plain English: roughly half of inbound calls don't reach a human, and the calls that do reach a human are where almost all the revenue lives. The voicemail is not a near-miss; it is the loss.
The after-hours version of this is worse. A burst-pipe call at 11:30pm is a $400-$1,500 job for a plumber. The homeowner is not patient. They will dial three numbers in the first ten minutes, and the first one to answer with a human voice and a time gets the job — that is just how panic works. The same call rolling to voicemail at your shop means it becomes the next plumber's $1,200 emergency. And the homeowner does not call back to apologize.
The daytime version is quieter but adds up. Your dispatcher is on the phone with a real job. The line rings. The caller has a question your dispatcher has answered three hundred times — do you service my ZIP, do you do tankless, what's the diagnostic fee, are you open Saturdays. If the dispatcher does not pick up in 30 seconds, the caller is on to the next listing. The 27% miss rate Housecall Pro reports is mostly made of those calls. None of them feel like a big loss in the moment. All of them add up.
There is also the triage problem most contractors barely talk about. Not every after-hours call is an emergency. A leaky faucet at 11pm is annoying but it can wait until morning. A no-hot-water call in July is annoying but it is not the same as no-hot-water in February. A burst pipe is an emergency. A breaker tripping repeatedly while the homeowner can smell something hot is an emergency. Right now, most home services businesses either route everything to the on-call tech (which burns out the tech and rolls trucks at midnight for jobs that could have waited) or route nothing to the on-call tech (and lose the real emergencies to a competitor). Neither is the right answer.
And underneath all of this sits the cost-of-time question. The U.S. Bureau of Labor Statistics reports the May 2024 median annual wage for plumbers, pipefitters, and steamfitters at $62,970, with roughly 44,000 openings projected each year through 2034. Every hour a dispatcher spends answering "what's your service-call fee" instead of routing a $4,500 furnace replacement is roughly $30 in fully-loaded labor doing the lowest-value work in the building.
What changes for your business
An AI chat assistant tuned for home services fixes three problems at once: it captures the after-hours calls that would otherwise go to voicemail, it answers the daytime FAQ load so your dispatcher can stay on revenue-generating calls, and it triages the difference between a real emergency and something that can wait until morning. None of this requires changing the field service software your team already knows.
The assistant lives in a chat bubble on your website, available every minute the site is up. When a visitor opens it, they get a short friendly greeting in your company's voice and an offer to help. The first question the assistant asks separates the routine from the urgent — "is this an emergency, a quote, a question, or scheduling a service call?" — and the conversation branches from there.
Triage is the load-bearing piece for home services and it is where most generic chat bots fall over. The triage logic is custom to your trade and your on-call policy, written in plain rules during the build. A plumbing build might route a burst pipe, a sewage backup, a no-water situation in freezing weather, or an active leak the homeowner cannot stop straight to the on-call tech via SMS — with name, address, callback number, and a one-line situation summary pre-filled. The same build would route a slow drip, a running toilet, a clogged drain that's not backing up, or a "we'd like a quote for a tankless" to the next-business-day schedule, booked directly through your field service software. An electrical build draws the line at smells of burning, sparks, or repeated breaker trips with hot panels. An HVAC build adjusts based on season — no-cooling in a July heat wave is treated differently than no-cooling in October.
For the routine load — the daytime FAQ questions your dispatcher has memorized — the assistant pulls answers from a scope you define in writing during the build. Your service area as a ZIP or town list. Your published flat-rate pricing and service-call fee, with anything diagnostic or parts-dependent explicitly off-limits. Your business hours and weekend coverage. Your specialties (tankless, sewer line, panel upgrades, mini-split, attic insulation, whatever your trade actually does). What a first visit looks like. The answers come back in seconds, in your company's voice, and the homeowner either books the call through your scheduling tool or leaves a request that lands in your dispatcher's morning queue with the question and context already captured.
The handoff to your field service software is the unglamorous piece that makes the whole thing work. Whether you run ServiceTitan, Housecall Pro, Jobber, FieldEdge, or something else, we work alongside the platform — pushing booked jobs into the scheduler the way the platform supports, routing on-call SMS through your existing emergency line, and not asking your dispatcher to log into a second dashboard to see what came in overnight. The morning queue is just there, in the system they already use, sorted and triaged.
For your business, what changes is concrete. The 11pm burst-pipe call gets your tech on the phone in three minutes instead of rolling to voicemail. The 11pm leaky-faucet call gets scheduled for 8am and the on-call tech keeps sleeping. The daytime FAQ load comes off the dispatcher's plate so they can route the actually-busy hours efficiently. The estimate request that lands at 2am on Saturday is in the dispatcher's morning queue, with the homeowner already acknowledged and the visit type pre-tagged. ServiceTitan's research shows 47% of larger contractors get 11-15% of their income from estimate follow-up — the assistant is the first step in not letting that follow-up start cold.
AI Chat Assistant for Plumbers, Electricians, and HVAC
A 24/7 chat assistant on your website that triages midnight emergencies the way a smart dispatcher would, answers the routine service-area and pricing questions that bury your phone all day, and hands the actually-urgent calls to your on-call tech with the situation pre-summarized — so the burst pipe at 11pm doesn't become the next plumber's job.
What we build for home services
A first-phase deployment ships in 2 to 3 weeks and lands as a working assistant your team does not have to think about after week three.
What you get when the build is done: a chat bubble on every page of your website, branded in your company's look and voice, available every minute the site is up. A defined scope of routine questions the assistant can answer — your service area as a ZIP or town list, your published flat-rate pricing and service-call fee, your hours and weekend coverage, your trade specialties, your insurance and warranty basics, what a first visit looks like — written down and reviewed with your team before go-live.
Custom emergency triage logic tuned to your trade. For plumbing, the burst pipe / sewage backup / no-water-in-freezing-weather / active-leak rules and the slow-drip / clog / quote / next-day rules, written explicitly. For electrical, the burning-smell / sparks / repeated-breaker-trips-with-hot-panel rules and the dim-bulb / outlet-not-working / quote rules. For HVAC, the season-aware no-heat-in-winter and no-cool-in-heat-wave rules versus the maintenance-and-quote routing. The line is drawn during the build, in writing, with your team's input.
A clean handoff into the field service software you already run. Booked jobs land in your scheduler the way the platform supports — through API, through an existing communication add-on, or through a thin middle layer if that is the cleanest path. We confirm what is reachable on the first call before quoting integration scope, so there are no surprises. On-call SMS routing goes through your existing emergency line, with the situation summary pre-filled so the tech reads the context before calling back.
A short list of guardrails written in plain English. The assistant is allowed to share published flat-rate pricing and your service-call fee. It is not allowed to quote diagnostic work, parts-dependent jobs, or anything custom — those route to a callback or a scheduled estimate. It is allowed to confirm service area inside your ZIP list. It is not allowed to promise service outside that list. It is upfront about being automated and offers a human handoff on every screen.
A simple weekly report your owner or operations lead can scan in a minute. How many conversations the assistant handled, how many emergencies it routed and where, how many jobs booked themselves through the scheduler, how many contacts handed off to your dispatcher, where the assistant said "I don't know" so the scope can keep getting sharper.
You stay in control of the voice, the scope, and the on-call policy. We do the building, the wiring, the testing, and the tuning over the first 30-60 days using actual transcripts from your homeowners. After it is live, the only thing your dispatcher has to do is pick up the conversation when a handoff lands — and answer fewer of the calls they were tired of answering anyway.
Outcomes you should expect
What this delivers
- Catch the 11pm burst-pipe call your dispatcher would have missed — and route the actually-urgent ones to the on-call tech while non-urgent calls book themselves into the morning schedule.
- Free your dispatcher from answering the same five questions (service area, ballpark price, do you do tankless, do you work weekends) so they can focus on routing trucks and closing warm quotes.
- Triage true emergencies from things that can wait — a burst pipe routes to on-call, a slow drip schedules for tomorrow, a no-hot-water in February routes to on-call, a no-hot-water in July schedules for tomorrow.
- Capture the after-hours estimate request that would otherwise end up at the next plumber on the search results page — research shows the first business to actually answer wins the job.
- Hand off cleanly to your field service software (ServiceTitan, Housecall Pro, Jobber, FieldEdge, or whatever you run) so the booked job lands where your dispatcher already lives, not in a separate inbox.
Illustrative scenario
What this typically looks like
The scenario below is illustrative — a representative outcome for a business that fits this service profile, not a claimed client engagement.
This is an illustrative scenario, not a description of a specific client engagement. The numbers are typical ranges, not a guarantee.
Picture a four-truck residential plumbing and drain company in a suburban metro market. They run on a major field service platform, have one full-time dispatcher and a part-time after-hours answering service that charges per minute. The baseline pattern: roughly 25-30% of inbound calls go to voicemail in any given week, mostly during the lunch crunch and after 6pm. The owner takes the 11pm calls personally about three nights a week — sometimes a real emergency, sometimes a leaky faucet that could have waited until morning. The dispatcher spends maybe 30% of her phone time on FAQ questions that don't end in a booked job. The answering service costs $600-$900 a month and only catches the calls that come through the main line, not the website.
After installing the AI chat assistant with triage tuned to plumbing emergencies, the pattern typically shifts over the first 60-90 days. The website captures 15-25 after-hours conversations a week that previously would have been form submissions sitting in an inbox until Monday. Of those, maybe 2-4 a week trigger the true-emergency route and text the on-call tech — and historically those are the high-ticket jobs that would have gone to a competitor. Another 10-15 schedule themselves directly into the next-business-day calendar through the field service software. The remainder leave contact info for a callback in the morning, already pre-tagged with what they need.
On the daytime side, the dispatcher typically gets back 2-4 hours a week from FAQ deflection — questions about service area, weekend coverage, ballpark pricing for common jobs, and "do you work on tankless." That time goes into routing trucks more tightly and following up on estimates that would have gone cold. The owner stops getting woken up for the leaky-faucet calls because the triage handles those. The answering service either gets replaced by the chat or gets demoted to backup for the main phone line.
The actual dollar lift will vary by market, season, trade mix, and how cleanly the triage scope is drawn. The shape of the math does not. The phone stops bleeding and the dispatcher stops being a switchboard operator.
Common questions
What buyers ask before reaching out
What makes an AI chat assistant for home services different from a generic chat bot?
A generic chat bot answers FAQs and collects emails. A home-services-tuned assistant has to do those things plus triage emergencies — and emergency triage is the load-bearing piece. A burst pipe at 11pm needs to route to your on-call tech. A leaky faucet at 11pm needs to schedule for the next business day. A gas smell needs an immediate 'call 911 and your utility' response before anything else. The assistant we build for plumbing, electrical, and HVAC contractors knows the difference, asks the two or three triage questions that draw the line, and routes accordingly. That logic is custom to your trade and your on-call policy.
How does the after-hours triage actually work?
When someone hits the website chat at 11pm, the assistant asks two or three short questions — what's happening, is water actively running or the breaker actively tripping, is anyone in danger. Based on the answers, it does one of three things. True emergency: it captures the name, address, phone, and a short situation summary, and texts your on-call tech immediately with everything pre-filled. Maybe-emergency: it offers a callback within 30 minutes from your on-call line, with the homeowner's permission. Not-an-emergency: it offers the first available next-day slot and books it through your field service software. The homeowner is not stuck explaining the situation twice.
Will it work with ServiceTitan, Housecall Pro, Jobber, or FieldEdge?
We work alongside your field service software rather than replacing it. The exact depth of the connection depends on what your platform exposes — some allow a direct booking handoff through their API, others work better through their existing communication add-ons, and some need a thin middle layer. We won't promise a deep two-way integration we haven't actually built. On the first 15-minute call we look at your stack and tell you what is reachable, what is not, and what the cleanest handoff looks like for your setup.
What stops the assistant from quoting a price it shouldn't quote?
Strict scope, set in writing during the build. The assistant is allowed to share a service-call fee and any flat-rate pricing you publish on your site. It is not allowed to quote anything diagnostic, anything that depends on parts, or anything custom. For those, it says clearly that a tech needs to see it before a real number is possible, and offers either a same-day callback or a scheduled estimate. The fail mode is a bot that bluffs a price; the system we build is one that says 'I don't know that one — let me get someone who does.'
Can it actually answer service-area questions, or will it tell people in the next county we'll come out?
Service-area is one of the things we lock down hardest. The assistant works from a ZIP code list or a town list you give us, and asks for the homeowner's ZIP before promising service. If the ZIP is in your area, it confirms and proceeds. If it's outside, it says so plainly — and depending on your preference, either offers a referral to a partner you trust or just thanks them and closes the conversation cleanly. No promises that turn into wasted truck rolls.
What happens during a real emergency — is the homeowner stuck talking to a bot while their basement floods?
No. As soon as the triage flags 'true emergency,' the assistant stops asking questions, confirms the address and a callback number, and pushes the situation directly to your on-call tech via SMS — with a one-line summary like 'Burst pipe, basement, water shut off unknown, address X, callback Y, time 11:47pm.' The homeowner gets a confirmation that someone is on the way, plus a short list of stop-the-bleeding steps (where the main water shutoff usually is, how to kill power to a wet area). The bot doesn't try to handle the emergency itself — it gets a human on the line faster than voicemail would.
How quickly can this be live on our website?
A first cut is usually live in 2 to 3 weeks from kickoff. Week one is gathering your service area, your published pricing, your on-call policy, your trade specifics, and writing the triage logic. Week two is wiring the booking handoff into your field service software and the on-call SMS routing. Week three is testing with real visitors and tuning. The first version is deliberately rough on the edges — it handles the obvious 80% reliably and routes everything ambiguous to your team. We tune it over the following 30-60 days based on actual transcripts.
What does this cost compared to a 24/7 answering service?
A 24/7 live answering service for home services typically runs $300-$1,500+ a month depending on call volume, with per-minute charges that scale fast in a busy season. An AI chat assistant build with us is a one-time setup in the $3,000-$6,000 range plus a monthly platform cost of $50-$300 depending on conversation volume — no per-call charges that punish you for a busy week. Most contractors recoup the setup in the first one or two emergency calls captured that would otherwise have rolled to voicemail.
Ready to see what this looks like for your business?
A free 15-minute call. We talk about your business, the time and revenue you'd unlock with the right automation, and what the first 30 days could look like.