Why Aurora HVAC crews drive to houses with nobody home

It is 1:10 on a Tuesday and your tech is parked in front of a house on the west side of Aurora. The ticket says one to four, no cool, second visit. He knocks and nobody answers. The number on the work order rings into a voicemail box belonging to a man who is at work until five.

So he calls the office. The dispatcher is on the other line with a customer asking where her tech is, so he waits in the truck. At 1:25 he leaves a door tag and drives twenty minutes to the next call, which was scheduled for three and is not ready for him either. The homeowner calls at 5:40 that evening, annoyed, and gets rescheduled to Thursday. He had already waited home on Monday for the first visit.

Two hours of a technician's day are gone and the customer thinks you stood him up.

What a failed roll actually costs

You sell a technician's hours in a truck, and there are only so many of them in a day. A visit where nobody answers the door costs more than the appointment. It costs the drive out, the wait, the drive back into the rotation, the slot another customer could have had, and then the second visit on Thursday that nobody is paying extra for.

People are not skipping these on purpose. A three hour window asks a customer to give up an afternoon on a guess. If nothing confirms the day before, and nothing tells them the truck is actually coming, the sensible move on their end is to run an errand and hope. Some of them get back in time.

The confirmation call is supposed to prevent all of this, and it is the first thing that gets dropped. Whoever would make it is the same person answering the phone, ordering a part, and explaining to the 8 a.m. why the first bill was what it was. A confirmation call has no deadline of its own, so in a busy week it does not happen, and the busy weeks are the ones with the fullest board.

Then there is rescheduling. A customer who cannot make Wednesday has to call during business hours, wait for whoever holds the schedule, and commit to a new window on the spot. Plenty of them never get around to it, and you find out when the truck is in the driveway.

None of it gets counted either. Nobody can say how many rolls failed last month or why. The dispatcher remembers a rough week. Whether it was bad phone numbers, windows too wide to plan around, or the same four customers, nobody knows, so October goes the same way.

What replacing it looks like

Individually these are unremarkable. What matters is that they run whether or not anyone at the office has a free minute.

• The appointment confirms itself the moment it is booked, by text and email: the date, the window, the address on file, what the visit is for, and what it costs to come out.
• A confirmation goes out the afternoon before. Reply C to confirm, or tap a link to move it. Anything still unconfirmed by the end of the day lands on one short list, and the dispatcher works that list instead of calling everybody.
• The morning of, the customer gets a narrower window and the tech's name.
• When the tech closes out the previous job, the next customer gets an on the way text with a real arrival estimate. That one message brings a lot of people back to the house.
• At the door with no answer, a sequence runs on its own: a call, then a text with a one tap reply, then a timer. When the timer runs out the tech is released to the next job and the dispatcher watches it happen instead of hearing about it later.
• The freed slot gets offered to jobs already waiting in that part of town, so a hole in the afternoon has a chance to fill before it is gone.
• Maintenance agreement visits get scheduled on their own cycle instead of whenever somebody opens the list in the spring.
• Every failed visit gets logged with a reason, so at the end of the month there is a count and a cause.

I have built these mechanics in other shapes. A freight forwarding client's quote and inquiry requests used to arrive as loose email and now come in structured and tracked. A pool service client's leads get captured and followed from first contact through to the job. A service schedule runs on the same pattern: it gets recorded when it happens, the customer hears back without waiting on anyone, and a person is pulled in where there is a decision to make.

What stays human

Everything requiring judgment. Diagnosis, obviously. The call where a fifteen year old system is dead in August and somebody has to sit at a kitchen table and talk honestly about repair versus replace. The decision about who gets bumped when a no heat call comes in at 4:30 on the coldest day of the year, because that is really a question about which customer you can afford to disappoint, and no scheduling tool should be answering it. The system confirms, texts, logs, and refills. Your dispatcher gets the day back for the calls that need a person.

When it is not worth building

If you run one truck and call every customer yourself the night before, you already have the system and it works. This gets worth building when the schedule outgrows one person's attention, usually two or three trucks in, or when you notice you are running a second visit most weeks because nobody was home for the first.

When it is worth building, builds start at $500 one time, then a flat monthly from $99 after an included run-in period, cancellable. There is more on the HVAC page and on what I build for businesses in Aurora.

Where to start

The first step is a free 20 minute process audit. We walk back through last month's schedule, count the visits that failed and why, and put a number on the hours your trucks spent driving to empty driveways. If a system pays for itself, build it. If it does not, you keep the map either way. Start here.