Sort Kentucky from Indiana in the first thirty seconds of the call
The state a caller is standing in changes whether the job is yours at all, so it belongs before the symptom, not after the address.
Put it in the script as the second question, right after what is happening. Are you on the Kentucky side or over in Indiana. Callers answer that instantly, and it costs you nothing.
If the answer is Indiana and you do not work there, the call is not a failure. It is a referral, done fast and warmly. Have a name ready. People remember who helped them at eleven at night, and their neighbor in Clarksville is not your customer anyway.
If you do work both sides, the answer changes the routing, the crew and sometimes the arrival window. Getting it up front stops the truck from being assigned twice.
Log the state on every inquiry. Without it, your reporting averages two markets with different travel times and different economics into a single number that describes neither.
Six in the morning and ten at night are booking windows, not overflow
In a metro where a lot of people work nights, off hours availability converts because it is genuinely useful, not because it sounds good in an ad.
Look at when your inquiries actually arrive, hour by hour, and compare it against when someone is there to answer. Most shops find one or two hours with real volume and no coverage.
Offer appointment windows that fit shift work. An early morning arrival for someone who gets home at six, or an evening slot for someone who sleeps during the day, is worth more to that customer than a discount.
Answer rate is the metric to watch, and it should be tracked separately from close rate. A call that went to voicemail is not a lost sale yet, but it is a race, and you are behind.
Test a callback promise you can keep. If your after hours system tells someone they will hear from a human by seven in the morning, and you deliver, that alone will change how many of those inquiries survive the night.
Phase the Old Louisville repipe or the estimate sits until spring
Big old-house numbers do not stall on trust, they stall on the size of the check, and a phased scope is a conversion tool.
A whole-house repipe in a Germantown or Old Louisville home is a decision most owners are not ready to make in one sitting. Presenting only the full number gives them one way to say yes and many ways to delay.
Offer a staged path where it is technically sound. The failing branch first, then the rest on a schedule. Say clearly what does not get solved by stopping halfway, so nobody feels sold to.
Put the estimate in writing with photos of what you saw, a plain description of the work and what is excluded. Old-house owners share estimates with a spouse, sometimes with a contractor. A verbal number cannot survive that conversation without you in the room.
Set the follow-up before you leave. A specific day, agreed out loud, beats an open promise to check in. Then keep it, because most competitors will not.
Verbal driveway quotes are the leak nobody writes down
The number your technician says out loud at the truck never enters your system, so it never appears in any report you review.
Ask your techs how many times last month they gave a price on site and heard nothing back. The answer is usually higher than anyone in the office expects, and none of it is tracked.
Fix it with a habit rather than software. Every quoted number gets logged the same day with the address, the scope and the reason it did not book on the spot.
Give techs a way to send the customer something before they drive off. A photo and a written scope from a phone is enough, and it turns a conversation into a document the customer can act on.
Once those quotes are visible, you will find patterns. One service line that never closes on site. One neighborhood where the numbers go quiet. One tech who books everything and one who books nothing, which is a coaching problem, not a marketing problem.
Close rate is a technician number before it is a marketing number
Two people can take the same inquiry from the same ad and produce very different outcomes, and averaging them hides your best lever.
Break booking and close rates down by the person who handled the call and the person who ran the appointment. It is uncomfortable and it is usually the most valuable report in the business.
Listen to a handful of recorded calls a week, chosen from the ones that did not book. Not to catch anybody out. Patterns show up fast: talking price before scope, missing the state question, failing to offer the next available slot.
Score the intake on a short checklist instead of a vibe. Was the state established, was the water status asked, was a specific window offered, was a follow-up set.
Coaching one dispatcher usually moves more revenue than any change to a website button, and it moves it this month.
Measure the callback across a heating season, not a two week test
One metro does not generate the volume to prove a small change quickly, so change things that are big enough to see and measure them over a long enough window.
Split testing a headline or a button color needs sample sizes a single market rarely produces in a useful timeframe. Waiting for statistical proof means waiting past the season you were trying to fix.
Change the things whose effect is large and legible instead. Answering hours. Whether estimates are written or spoken. Whether a follow-up is scheduled. Whether the state question is asked. Those move numbers you can see without a calculator.
Run one change at a time and hold it long enough to survive a cold snap and a quiet week. Heating season here is long enough to give a fair read if you start early.
Keep a simple log of what changed and when. Six months later, when the schedule board is full and nobody remembers why, that log is the difference between repeating what worked and guessing again.
Questions we actually get
- What should we fix first if we can only do one thing?
- Answer rate, almost always. Pull a month of call records and count how many rang out, went to voicemail or were abandoned while on hold. If that number is meaningful, no website or ad change will outperform simply having someone pick up. It is also the fastest thing to measure again in thirty days.
- Is it worth turning down work from across the river?
- If you cannot legally or practically serve it, yes, and quickly. The cost of a slow no is a customer waiting on you and a dispatcher tied up. Handled well, with a referral, it also generates goodwill. Whether you can serve a given job in Indiana is a licensing question worth confirming with your own counsel rather than deciding case by case on the phone.
- How do we track conversions when most inquiries come by phone?
- Call tracking numbers by source, recordings, and an outcome code entered for every call. The codes matter more than the technology. Booked, quoted, out of area, wrong service, price only, no answer. Without them you can see how many calls arrived and nothing about which ones were worth having.
- Should we run A/B tests on our site?
- Usually not on small elements. One metro rarely produces the traffic to prove a small difference before the season ends. Larger structural changes, like adding online booking or publishing price ranges, produce effects big enough to read without a statistics package. Make one change, hold it, and compare against the same period rather than the previous week.
- How much of this is marketing and how much is operations?
- Most of it is operations, which is why it works. The inquiries you already generate are the cheapest source of growth you have, and the changes that convert them live in scripts, schedules, estimate formats and follow-up habits. We would rather fix that first and then spend on demand, because more inquiries into a leaky intake mostly produces more missed calls.
