A Clarksville ZIP and a Middletown ZIP cannot get the same answer on the phone
Which side of the Ohio River the caller is on decides whether the call is worth taking, and it should be the second thing anyone asks.
One metro across two states means every inquiry carries a licensing and scheduling question before it carries a technical one. Jeffersonville, New Albany and Clarksville sit in Indiana. Louisville Metro sits in Kentucky. Requirements differ by state, so confirm your own standing with the relevant agency and set the script accordingly.
Put the service address at the top of the script, not near the bottom with the email. Whoever answers should be able to say yes or no within twenty seconds, which respects the caller and protects the schedule.
If you serve only Kentucky, teach the phone to decline warmly and refer. A clean no takes thirty seconds. A soft maybe takes a dispatcher an hour and ends the same way.
Inside Kentucky, keep asking. Louisville Metro and Jefferson County are merged, while Oldham and Bullitt are not, and a run out past the Gene Snyder is a different day than a run to St. Matthews.
Cold snap overflow looks like a marketing problem and is really a staffing one
The days you lose the most work are the days the phone rings most, and almost nobody looks at that data afterward.
Pull the call log from the last hard freeze. Count the calls that rang out, the ones that hit voicemail, and the ones where hold time crossed two minutes. Every one of those was demand you had already paid for in some form.
Missed calls in this business cluster in predictable windows. Early morning after a freeze, the first hour after a shift change, and the evening rush when everybody gets home to a cold house. Staff to the pattern rather than to an average.
Overflow answering is worth arranging before the season, not during it. A service that can take an address, a system type and a callback window keeps the inquiry alive even when your own line is full.
Callbacks have a short shelf life in a no-heat situation. A returned call the next morning is generally competing against a contractor who already showed up.
Shift workers call at hours when only your voicemail is working
A logistics and air freight heavy metro produces households doing business at hours most contractors have written off.
Look at when your inquiries actually arrive rather than when you assume they do. Form submissions at one in the morning and calls at six are normal here, and they are usually somebody with limited daytime availability.
Test the after-hours experience personally. Call from a number your office does not recognize, at an hour nobody expects, and listen to what a stranger hears. Owners are routinely surprised, and the fix is often one script change.
Booking outside standard hours is a real differentiator in this market rather than a slogan, and it is worth stating precisely. Saying you answer until ten converts better than saying you are always here for you.
If you cannot staff those hours, at least make the overnight form set an honest expectation and confirm a morning callback time. Certainty beats availability more often than owners expect.
Most furnace quotes get decided three days later by somebody who was not home
Replacement is a household decision, and the person who signed nothing at the kitchen table is often the one who decides.
A proposal handed over and never followed up is the most common leak in an HVAC business. The technician did the work, the office did the paperwork, and nobody called back on day three.
Build a fixed follow-up sequence and put it on somebody's job description. A call, then a written recap the absent decision maker can actually read, then a final check before the quote expires. Timed, not remembered.
Write the proposal for the person who was not there. What failed, what the options are, what changes between them, what happens if they wait. Somebody reading it cold at the kitchen table should be able to follow it without a technician present.
Log why quotes were lost, and read the list monthly. Price, timing, a competitor, or nobody called back. The fourth reason is the one you can fix this week.
Tag every booked job to a county or your service area is a guess
Measurement in a two state metro means splitting the numbers before you average them.
Booking rate in Jefferson County and booking rate in Floyd County are separate results. Blend them and you learn nothing about either, and you cannot tell whether a soft month was a demand problem or a coverage problem.
Track the whole path. Inquiries received, inquiries answered live, appointments booked, appointments run, quotes issued, jobs sold, invoices collected. Most shops count the first and the last and guess about the middle.
Reconcile the dispatch board against lead source at least monthly. Marketing reports and job records disagree in almost every shop we have looked at, and the disagreement usually hides the best performing source.
Count the jobs you turned down and where they were. A month of declined addresses is the most honest map of your real service area anyone will ever hand you.
One metro is a small sample, so run experiments in months and not in sessions
There is not enough traffic here for small split tests to tell the truth, and weather makes it worse.
A single metro rarely produces the volume needed to detect a small change in a form or a button, and a cold week will swamp whatever signal there was.
Test things big enough to matter. A rewritten phone script against the old one. A form that asks for the address first against one that asks for a name. An after-hours path with a live answer against one without. Changes at that scale show up even in modest numbers.
Run them in month long blocks and compare like periods, ideally the same stretch of season year over year. Comparing January against April in this business is comparing two different companies.
Start by listening rather than testing. Twenty recorded calls will usually reveal more fixable problems than a quarter of experiments, and the fixes are free.
Questions we actually get
- Where do most HVAC inquiries actually get lost?
- In three places, usually. Unanswered calls during freeze weeks and shift change hours, appointments booked without enough information to run, and replacement quotes nobody followed up on after day one. All three are process problems rather than marketing problems, and all three are cheaper to fix than to advertise around.
- Should intake ask for the address before the name?
- In this metro, generally yes. The service address settles the county, the state and the drive in one answer, which tells you within seconds whether the job is yours to take. Asking it last means a dispatcher discovers the problem after investing five minutes in the call.
- How do we handle the surge after a hard freeze?
- Arrange overflow answering before the season starts, staff to the pattern in your own call log rather than to an average day, and make sure whoever answers can capture an address, a system type and a callback window. Callbacks in a no-heat situation have a very short shelf life.
- Is A/B testing worth it for a single metro HVAC company?
- Small tests, rarely. There is not enough traffic to separate a modest effect from a cold week. Bigger changes tested over month long blocks work better: a rewritten phone script, a reordered form, a live after-hours answer. Start with call recordings, which are free and usually more revealing.
- What should we measure besides lead count?
- The full path from inquiry received to invoice collected, split by county so the Kentucky and Indiana sides can be read separately. Add the jobs you declined and where they were. Lead totals flatter every marketing report ever written, and they hide exactly the leaks worth fixing.
