Where the revenue is actually leaking
We don't start from an industry benchmark for what a missed call is worth. Plenty of consultants will hand you a number pulled from a survey of businesses nothing like yours. We start from your own call log, your own average job value, and your own close rate, because those are the only numbers that tell you whether fixing something is worth real budget or just a distraction from the thing that's actually costing you more.
For most trades businesses, the leak sits in one of four places:
- Missed calls. The phone rings while a tech has both hands inside a furnace, and the caller doesn't leave a message — they call the next name on the search results.
- Slow quotes. A homeowner comparing three contractors usually books with whoever gets back to them first with a real number, not whoever eventually sends the most thorough one.
- Unsold estimates. A quote goes out, nobody follows up, and the job quietly goes to a competitor who called back.
- No follow-up. Maintenance plans lapse without a renewal reminder, and customers who are due for a repair never hear from you until they call someone else first.
Most businesses have all four to some degree. The point of this engagement is finding out which one is actually the biggest, in your business, this year — not which one gets talked about most on trade forums.
What this engagement actually produces
Three things, written down, that you can act on without us: a short list of where the leak is worst, sized against your own call log, quote log, and job numbers rather than a generic benchmark; a call on each item for whether it's worth building something custom or just turning on a feature you're already paying for; and a sequence — what gets fixed first so it pays for what comes next, rather than tackling the biggest idea first and running out of budget before it proves itself.
That's the whole deliverable. There's no governance framework, no committee sign-off process, no compliance documentation — that apparatus exists for industries where a mistake creates legal exposure. A home service business needs an answer, not a policy binder.
Whether you actually need this at all
Here's the honest version: if you run one or two trucks and you already know your problem is that nobody answers the phone after 5pm, you don't need a strategy engagement. You need the phone fixed. Go read what a voice AI system actually does and book a call to scope that directly — a strategy phase in that situation is an extra step that delays the fix without changing the answer.
This engagement earns its cost when there's real ambiguity to resolve: more than one candidate problem competing for the same budget, a field-service platform with AI features nobody's actually evaluated against what you need, or a business big enough — several trucks, multiple call types, a seasonal swing that triples call volume for a few months — that guessing wrong wastes real weeks. If that's not your situation, say so on the first call. We'll point you straight at the build instead of selling you a phase you don't need.
Build vs buy: what your field-service platform already does
Most owners we talk to already pay for a field-service platform — ServiceTitan, Housecall Pro, and Jobber are common examples across the trades, though which one you run doesn't change the question. A lot of these platforms now ship some form of AI feature: auto-suggested dispatch, an AI answering add-on, automated review requests. We are not partnered with any of them, and we don't assume you run any particular one.
The question worth answering before anything gets built is simple: does the feature you're already paying for actually close the gap you have? An AI answering add-on that only handles scheduling for existing customers and drops anything unfamiliar to voicemail doesn't solve an after-hours emergency problem, even though it looks like it should on the feature list. Sometimes the right move coming out of this engagement is to properly configure and start using what you already own. Sometimes the gap is real, and a purpose-built system earns its cost. Telling you which one applies, honestly, is most of what this engagement is for.
Walkthrough: sizing a missed-call problem from your own call log
A three-truck HVAC company comes to us convinced they need a full AI receptionist. Before recommending anything, we pull their actual call log for a recent stretch of weeks and count what went to voicemail or rang out unanswered, split by when it happened — during business hours, when someone should have picked up, versus nights and weekends, when nobody was scheduled to.
Their own close rate and average job value turn that count into what the gap is actually worth, not an industry figure. In one case like this, the finding cuts the project down rather than up: business-hours misses turn out to be rare because staffing already covers most of it, and the real leak is concentrated in the after-hours and weekend window. That's a narrower, cheaper build than the owner walked in assuming, and it's a better outcome than talking them into the bigger system they thought they needed.
Walkthrough: the estimate that goes out and never gets chased
A remodeling contractor sends quotes reliably and still isn't sure why the close rate feels low. We line up the quote log against the job log — how many estimates went out, how many became a booked job, and how long it took on the ones that closed versus the ones that didn't. The pattern that usually shows up is a batch of estimates, often the larger ones, that sit untouched for weeks with no second contact, and simply decay.
The fix here rarely needs a full system rebuild. It's often a narrower AI agent that runs a sensible follow-up cadence on exactly that batch of stalled estimates — which is a smaller, cheaper build than the "AI for everything" version the owner originally described wanting.
Walkthrough: what has to be true before dispatch automation helps
A landscaping company adding crews wants an AI system to assign jobs automatically. Dispatch today runs on a group text thread and a whiteboard photo, and two crews have shown up to the same address more than once. Automating that process as it stands would not fix it — it would just run the same confusion faster and with more confidence behind it, because a system acting on a schedule nobody trusts produces bad assignments quicker than a person would.
The honest output of the engagement in this case is not a build recommendation. It's a specific, smaller ask: get dispatch onto one reliable source of truth — often just using the scheduling board inside the platform they already pay for, consistently, for a few weeks — before any automation touches it. Sometimes the right answer from a strategy call is "come back to us once this part is fixed," and that's a better answer than building around a process that isn't ready to be automated yet.
What gets built next, and why the first project pays for the second
The sequence matters more than the size of any single idea. Whichever leak is both biggest and fastest to prove gets built first — for most trades businesses that's the phone, because a voice AI system that stops calls falling into voicemail is usually the single highest-leverage fix and the fastest one to see working.
What that first project recovers — jobs that would have gone to a competitor, hours an office person isn't spending on routine calls — is what funds and justifies whatever comes next. That might be an AI agent working the stalled-estimate queue or chasing lapsed maintenance renewals, or it might be workflow automation cleaning up the admin tail after a job is done — invoicing, reminders, review requests. Whichever it is, it inherits the sizing and sequencing from this engagement rather than starting from a blank page, and the wider picture of how all of this fits a trades business day to day is on our home services industry page. For businesses outside the trades, the same approach is covered on the general AI strategy page.