Why the phone keeps going to voicemail while your trucks are out on jobs, what's actually safe to hand off, and an honest answer on whether it can tell a burst pipe from a routine tune-up.
A 1-10 truck HVAC or plumbing company usually has one or two people answering the phone — often the owner or a single dispatcher who's also texting a tech, updating the schedule, or answering a second line. When a third call comes in, it goes to voicemail, and most callers with a broken furnace or a leaking pipe don't wait around — they call the next name on the search results page instead. A dispatch agent, software that answers the call, asks what's going on, and either books it or flags it as urgent for a person, can pick up every call the moment it rings, including at 11pm on a Sunday when your water heater customers' pipes are actually failing. Judgment calls — gas smells, ambiguous emergencies, price disputes — should still reach a person fast. This guide shows what's safe to automate first and what a real triage setup for this vertical actually looks like.
It's 7:40am. Your one dispatcher is on the phone confirming a 9am tune-up, and the office line rings again. She lets it go — she's mid-sentence with the last customer. It rings a third time twenty minutes later while she's texting a tech an address change. By 9am she has four voicemails, two of which are routine "can I book a filter change," and one of which is a family whose furnace died overnight and who, by the time anyone calls them back, has already booked the next company on their search results page.
This isn't a one-off bad morning. For a company running 1-10 trucks, the person answering the phone is almost always the same person doing dispatch, scheduling, and often billing — there's no dedicated call center, and there usually can't be one. Multiply four missed calls by every weekday, add in nights and weekends when your team is off the clock but pipes and furnaces don't check the calendar, and the number of calls that never get a live answer adds up fast.
A missed call doesn't show up as a lost job on any report — it just looks like a quiet day. But HVAC and plumbing work is disproportionately reactive: a customer with no heat or a flooding basement isn't going to leave a voicemail and wait patiently. They're calling because something is wrong right now, and if nobody picks up, they call the next result. Industry data backs up how much of this demand happens outside normal hours: CallRail's home services marketing report, citing Jobber booking data, found that 41% of jobs booked online for home service businesses come in after regular business hours — nights, weekends, and early mornings, exactly the windows a 1-10 truck shop is least staffed to answer live (CallRail, "31 home services marketing statistics every contractor needs to know"). If your phone coverage stops at 5pm or whenever your dispatcher's other line is busy, you're structurally missing a big share of the calls that matter most.
Hiring your way out of this — a second dispatcher, an answering service, after-hours coverage — is a real fixed cost for work that arrives in unpredictable bursts. A cold snap or a string of frozen pipes can double your call volume for three days and then go quiet. Staffing for the peak means paying for idle time the rest of the month; staffing for the average means getting buried every time the weather turns.
Before you touch any software, find out what's actually happening on your phone line. This takes about a week and no new tools:
| Call type | Safe to automate? | Why |
|---|---|---|
| Hours, service area, "do you do tankless water heaters" | Yes | Same factual answer every time, zero judgment needed |
| Scheduling a routine visit (tune-up, inspection, filter change, drain cleaning) | Yes | Checks the dispatch board and books the next open slot, same as a person would |
| Status check on an existing job ("is the tech still coming today") | Yes | The answer already lives in your dispatch system — it just needs looking up |
| Clear emergency: "no heat," "water pouring in," "pipe burst" | Yes, for triage — books/flags urgent and alerts on-call tech immediately | Keywords are unambiguous; speed matters more than a human picking up first |
| Ambiguous or ongoing issue ("it's making a weird noise," "not sure what's wrong") | No — route to a person | Needs judgment about severity that a script shouldn't guess at |
| Gas smell or suspected gas leak | No — immediate human/emergency-line escalation, always | Life-safety issue; never left to software alone |
| Price quote or billing dispute | No — route to owner or office manager | Pricing judgment and customer trust matter too much to script |
A dispatch agent is software that answers your phone, listens to what the caller says in plain speech, and decides what to do with it — book a routine slot on your schedule, look up an existing job's status, flag the call as urgent and alert an on-call tech, or transfer to a person if it isn't confident. It's not a phone tree where callers press 1 for service; they just talk the way they would to your dispatcher, and the system does what your dispatcher would do with the same information. In other words: it's a stand-in for the parts of the phone call that are pattern-matching and lookup, not judgment.
The sorting itself is often called triage — a term borrowed from emergency medicine, where patients are sorted by how urgent their condition is before anyone treats them. Here it means the same thing applied to calls: is this "book it whenever," or "someone needs to know about this in the next five minutes." A dispatch agent applies simple, clear rules to make that call — specific words like "no heat," "flooding," or "gas smell" route straight to urgent; everything else defaults to a normal booking flow.
Take the free 2-minute readiness assessment
Say a 4-truck HVAC and plumbing company takes about 25 inbound calls on a normal weekday. With one dispatcher juggling scheduling, texts to techs, and the phone, roughly two-thirds get answered live — the rest, somewhere around 8 calls, go to voicemail. If even 2 of those 8 would have booked a job — a reasonable estimate for calls where the customer actually needed something now — and the average ticket across routine and emergency work runs about $380, that's roughly $760 a day in jobs that never got booked. Over a 5-day week, that's about $3,800. Over a month, somewhere around $15,000. (Illustrative numbers for a mid-size HVAC/plumbing company — your real figure depends on your call volume, average ticket, and how many missed calls would have actually converted, which is exactly what the audit above is for.)
"I used to think the missed calls were mostly people who'd call back later if it mattered," is roughly how one plumbing company owner in this position might put it. "Then I actually looked at the voicemail log for a week and realized half of them were people with an actual leak, and none of them called back — they just booked whoever answered." That's an illustrative, representative way an owner in this spot tends to describe the problem once they look — not a specific named individual's quote.
This is the fair question, and it deserves a straight answer, not reassurance dressed up as one. The real concern isn't whether a caller minds a slightly different voice on the line — it's whether a system will correctly recognize a burst pipe or a no-heat call in freezing weather as urgent, versus treating it like a routine booking and losing hours a customer doesn't have.
For clear cases, this works well: specific phrases like "water is pouring out of my ceiling," "I smell gas," or "no heat and it's 20 degrees out" are unambiguous, and a well-built dispatch agent is set up to flag those immediately and get a tech or a person on the line fast — faster, in practice, than waiting through a busy signal. For genuinely ambiguous cases — a vague noise, a symptom the caller can't describe well — the system should be built to escalate rather than guess, meaning it treats uncertainty as a reason to involve a person, not a reason to make its best guess and move on. Gas leaks and other life-safety situations should always have a guaranteed fast path to a live human, as a fixed rule, not a probability the software calculates. And in the first several weeks of any setup, a person should be reviewing how the system is actually triaging real calls from your business — the same way you'd listen in on a new dispatcher's first week rather than trust them unsupervised on day one. If a vendor tells you their system needs zero human review from day one on live emergency calls, that's a sales claim, not a safe one, and it's worth walking away from.
This isn't "put AI on your phone" as a generic pitch. Unmanually starts by mapping your actual call patterns — what counts as urgent for your specific business (a slab leak is not the same urgency as a dripping outdoor spigot), how your dispatch board and tech schedules actually work, and where the line to "get a person on the phone now" needs to sit before anything goes live. We build the triage rules around your business, not a one-size-fits-all script, and we stay involved from onboarding through ongoing use — if a call gets mis-triaged, that's a real conversation with us to fix the rule, not a support ticket that disappears.
The direct payoff is the jobs you stop losing to voicemail — real booked revenue instead of a caller who moved on to a competitor while your line rang out. The bigger payoff is capacity: a shop that answers every call the moment it comes in, day or night, can book more of the jobs its marketing is already generating, without adding a second dispatcher's salary. And it gives the owner something rarer in this trade — an actual night or weekend off, without every emergency call quietly turning into a lost customer while you're at your kid's game.
Not ready to commit to anything yet? That's completely fine — leave your email on our presale waitlist and we'll let you know as soon as HVAC and plumbing dispatch and call-triage automation is live, including founding-member presale pricing before it opens to everyone else this October.
Don't automate your whole phone line on day one. Run the 3-step audit above first, then start with whichever call type is costing you the most — for most HVAC and plumbing shops, that's after-hours emergency calls, since those are the ones most likely to walk to a competitor if nobody answers. Our guide on the first 5 processes most small businesses should automate has more on sequencing this correctly across any business type.
The full pillar guide this article belongs to.
More on the review/spot-check pattern mentioned above.
The same phone-audit approach, applied to a different local business.
Track call types for a week, then split them into repeatable calls (hours, routine scheduling, status checks) and judgment calls (ambiguous issues, price disputes, gas leaks). A dispatch agent can answer and triage the repeatable and clearly-urgent calls immediately so nothing goes to voicemail by default.
For clear cases — "water pouring in," "no heat," "gas smell" — yes, and it should route those immediately. For ambiguous cases, a well-set-up system escalates to a person instead of guessing. Gas leaks and other safety issues should always have a guaranteed fast path to a human, and a person should review real calls in the first few weeks of any setup.
Take our free 2-minute readiness assessment — it walks through this same audit against your actual call volume and tells you honestly what to fix first.
Take the 2-min readiness assessment