When Your Phones Catch Fire for Six Weeks: A Survival Plan for Seasonal Call Spikes

The first heat wave of the summer used to wreck us. I ran operations for a home services group, and the day the temperature crossed ninety, every AC unit in a fifty mile radius decided to die at the same hour. The phone rang off the hook from seven in the morning. By noon we had three people on the desk who normally handled one, and we were still dumping callers into voicemail. I remember pulling the voicemail log the next morning and counting forty seven messages. Maybe a third of those people had already called the next shop on their list and booked.

That is the seasonal trap. The demand is real, it is predictable to the week, and it still beats you because you cannot hire and fire a receptionist for a six week window. So you either carry too much staff all year and bleed payroll, or you staff lean and watch your busy season leak revenue out the bottom.

I have lived both versions. Here is how I would handle it now.

Forecast from your own history, not your gut

Every seasonal business has a shape to its year, and most owners can describe it in vague terms. "We get slammed in the spring." Fine. But slammed when, exactly, and by how much? The fix is boring and it works: pull your call records and your booking records for the last two years and lay them out by week.

You are looking for three things.

• The ramp. When does volume start climbing off the baseline? It is almost always earlier than people think. For the AC business, calls started rising a good ten days before the first real heat, because people were testing units that had sat idle all winter.
• The peak. Which two or three weeks are the worst, and how much higher than baseline are they? In my experience the peak is usually two to four times your normal week, not the ten times it feels like when you are in it.
• The tail. How fast does it drop off? This matters because the tail is where overstaffing quietly burns money. The rush ends and the temp you hired is still sitting there.

Once you have the shape, you can plan capacity instead of reacting to it. A dental practice I worked with had a January spike every year the second insurance benefits reset. Completely predictable. We knew it was coming and we still got buried, because knowing is not the same as having coverage ready.

The math on overstaffing nobody wants to do

Here is the part owners avoid. Say your peak needs three people on phones but your normal day needs one. If you keep all three on payroll year round to be safe, you are paying two extra salaries for maybe eight weeks of real use. At even a modest wage with payroll taxes, that is real money sitting idle for ten months.

The honest alternative most shops reach for is seasonal temps. I have hired plenty. The trouble with a temp on the phones is that the phone is the worst job to learn fast. They do not know your pricing, they do not know which jobs you actually do, and they put callers on hold to go ask someone. During a spike, the person they need to ask is also drowning. So your overflow help slows down the people who were already keeping up.

Overflow coverage that does not need a chair

This is where the thinking has changed for me. The question is not "how many people do I hire for the busy season." The question is "what happens to call number two when call number one is already being handled."

Most phone systems just queue that second caller or send them to voicemail. And a caller on hold during your busy season is a caller dialing your competitor on their other phone. I have watched it happen on the floor.

The better answer is overflow that picks up instantly and actually handles the call. That is the case I would make for an AI receptionist as your overflow layer. LastWorker answers the phone, the website chat, texts, and email around the clock, and it does not care if forty calls land in the same ten minutes. Each one gets answered in under a second, in a voice that sounds human, in any of 97 languages. It already knows your hours, your pricing, your services, and your policies because you taught it that in a setup conversation that takes about fifteen minutes. No code, no new hardware.

What it does on a spike call is the same work your best front desk person does on a calm Tuesday. It answers the question, books or reschedules the appointment, captures the lead with name and callback number, takes a message, and when something genuinely needs a person, it transfers or escalates. The difference is it does ten of those at once without panicking.

Scale up and down without the hire and fire cycle

The reason seasonal businesses overstaff is that human capacity comes in fixed lumps. You cannot hire 1.4 receptionists. AI capacity does not come in lumps, which changes the whole calculation.

Think of it in two modes:

You are not turning anything on or off. The same system runs all year and the work simply shifts toward it when volume climbs. When the tail comes and things quiet down, your team takes the front line back. Nobody got hired for six weeks and nobody got let go after.

The pricing model is what makes this honest for a seasonal business. There is no monthly seat you are paying for in your dead months. You load a prepaid balance and pay per conversation it actually handles. Voice is billed per second at five cents a minute, chat and SMS per message, email per resolved ticket. In your slow season you spend almost nothing. In your peak you spend more because you are handling more calls, which is exactly when you can afford to, because those calls are booked jobs. A dedicated number is a dollar a month if you want one, and you can set auto-reload so the balance never runs dry mid-rush. The full pricing is there if you want to run your own numbers.

A few hard-won rules

Plan your peak coverage before the ramp, not during it. The week you realize you need help is already too late to set anything up calmly.

Decide your escalation rules in advance. Be specific about what gets passed to a human: a complaint, a custom quote, an angry caller, a job type you do not want booked without review. Everything else should resolve without a person touching it.

Watch your tail and pull human hours back as volume drops. The waste hides there.

And measure your missed-call rate before and after, not just your answered calls. The number that mattered most to me was never how many calls we took. It was how many we let ring out. That forty seven message morning is the one I think about. Every one of those was a person who needed help right then, and most of them found it somewhere else. A spike you cannot answer is not a busy season. It is your competitor's busy season, paid for with your marketing budget.