The two weeks that decide your HVAC season

You know the week before it happens. The forecast shows five straight days over 95, or the first real cold front is stacked up over the Ohio Valley heading for DC, and you can feel your stomach do the thing it does every year: here we go. Every AC unit in the county that's been limping along on a dying capacitor since May is about to finish the job. Every furnace that was "making a noise" all last winter is about to stop lighting entirely. And it's all going to happen in the same 72 hours, to everyone, at once.

This is not a slow ramp. It's a cliff. A DMV HVAC shop can run a normal week at 20-25 calls and then hit 60+ in the first three days of a heat wave — the phone doesn't ring more politely just because you're already on your fourth no-cool call of the morning. I want to walk through what that surge actually does to your call math, why the answering services built to catch it are priced to hurt you worst in exactly that window, and what your real options are. Fair warning: I sell one of them. I'll give you the honest math on the others too.

Why the surge is worse than it looks

Say your shop normally fields 20 calls a week and misses something like a quarter of them — a tech's mid-install, you're driving, you're elbows-deep in a condenser coil with the meter running. That's roughly 5 missed calls in a slow week. Not nothing, but manageable.

Now the heat index hits 105 for the first time this year. Call volume doesn't creep up 20% — it can triple or quadruple for three or four days as an entire service area's AC units fail together, because they're all the same age, installed in the same building boom, wheezing under the same heat. If your miss rate holds steady at a quarter (and it probably gets worse, not better, because you're now behind on every job and never near your phone), you're not missing 5 calls anymore. You're missing 15-20 in a single week — and a no-cool call in a heat wave isn't a $150 tune-up, it's an emergency-rate service call or a same-week install, worth several hundred dollars each. Do that math once — even conservatively, 15 missed calls at a 35% close rate and a $450 average job — and you're looking at over $2,300 in a single week quietly going to whoever else was answering their phone.

Here's the part that should bother you more: this is also the week you're least equipped to answer. You're not calmly available between jobs. You're running from a Bethesda no-cool call to a Manassas no-cool call to a compressor that seized in Alexandria, and the voicemail box is filling up with people who are, understandably, not thrilled about 95 degrees indoors with a toddler.

The voicemail you actually get back to is still homework

Say you do get to some of those voicemails, that night, after the trucks are back in the lot. That's not the call handled — that's the call rescheduled onto your evening. You listen to it, you call back, they don't pick up because they're also melting and also on the phone with someone else, you try again, you finally connect, you check a schedule that's already packed solid through Thursday, and you tell them the earliest slot is tomorrow afternoon — which, if a competitor's truck showed up that morning, is a slot they no longer need. The voicemail wasn't a missed opportunity you dodged. It was unpaid dispatch work, done at 9pm, for a job you might not even get to keep.

Why the meter runs hottest exactly now

Here's the part worth actually sitting with, whatever service you're paying for. Companies like AnswerForce price per minute or per call, with a plan cap and an overage rate once you're past it. That structure works fine in a normal week. It falls apart in a surge week, because a surge week is precisely when your call volume — and therefore your bill — spikes hardest. The metering doesn't know or care that this is the one week where the calls are worth the most to you; it just charges you more for more of them. Your highest-revenue week of the year and your highest-cost answering-service bill of the year land on the same seven days, every single time. That's not a coincidence, it's how the pricing model is built.

Your surge-coverage options, honestly compared

Do nothing, let it ring. Free, and it's what most solo HVAC shops default to, not because it's smart but because nobody's put a number on it. The cost is the $2,000+ a surge week can cost you, paid in full, silently.

Answer everything yourself, between jobs. You already do this to some degree. It doesn't scale past a certain call volume, and every minute spent triaging a call in a driveway is a minute not spent finishing the job you're actually on-site for — which pushes your whole day later, which pushes your voicemail pile bigger.

A per-minute human answering service. Real coverage, a real person, and — as covered above — a bill that spikes in lockstep with the surge that's supposed to be your best week of the year. If your call volume is genuinely low and steady, this can pencil out fine. If your business lives and dies by two weather events a year, the metering works against you at the worst possible time.

Hire seasonal help. Some shops bring on a part-timer for the two surge windows specifically. It can work if you already have someone reliable lined up, but hiring, training, and scheduling around two unpredictable weather windows a year is its own kind of scramble, and the timing is never exact — weather doesn't respect a start date.

A flat-rate AI receptionist. This is the category Dialkeep's in, so read the next bit with that in mind. The pitch built specifically for this problem: $199 a month, flat — the same bill in a 20-call week and a 90-call one. It picks up instead of ringing out, asks the address and the actual symptom — no cool air at all versus a system that's just running warm — and quotes your configured diagnostic fee exactly as you set it, the same on a 98-degree Thursday as on a slow morning in April, with no improvising a number to a stranger. The heat-wave week that breaks a metered plan's bill doesn't touch this one.

To be straight about the limits, and the big one first: it doesn't book. Booking isn't built, and neither is any route — automatic or manual — that carries what the caller told it back to you once they hang up. So look again at that 9pm pile two sections up, and be clear-eyed about the trade: a caller it answers doesn't leave you a voicemail either, which means on your side of the call you can end up with less to work with than voicemail gives you today, not more. It also doesn't diagnose a failed compressor or quote a repair over the phone, and it shouldn't — nothing should, at the phone-answering stage. What it actually buys you is the other end of the call: the homeowner sweating through hour six with a toddler gets a live voice, the right questions in the right order, and your real diagnostic fee, instead of four rings and a beep while you're on a roof in Manassas. That's a narrower claim than "surge coverage," and it's the one that's true. If you want to see whether it holds up during an actual surge scenario, the demo line is coming — when it's live you'll be able to describe a dead condenser on the hottest day of the year and judge it yourself.

Match the fix to your actual season

If your call volume barely moves between January and July, most of this doesn't apply to you, and a per-minute service or just answering it yourself might be entirely reasonable. But if your business genuinely has two weeks a year that make the difference between a good season and a break-even one, the fix you pick should be flat-rate by design, not despite the pricing model. Whichever fix you land on, keep the diagnostic fee itself consistent no matter who's fielding the call — here's a phone script that keeps it the same every time, surge week or not. For the fuller math on what a missed call costs a shop your size the other fifty weeks of the year, see how many calls a small shop actually misses, and if you want the flat number instead of a range, it's on the pricing page.