Is round-the-clock coverage worth paying for if you only get two night calls a week?

You pulled up the phone log for last month and counted what came in after you'd locked up. Weeknights, Saturdays, a Sunday. It came to about two a week. Now somebody's quoting you a monthly price to answer them, and the question in your head is fair: is anybody worth paying to catch two calls?

Here's the short version. Sometimes it isn't, and two a week is right where the answer can go either way. Paid coverage only earns the night jobs you aren't already winning some other way, so the test is one line: the booked jobs your night calls would produce, minus the jobs you already win from them, set against the quote divided by the gross profit on one of those jobs.

At Invoca's 2026 HVAC rates, two night calls a week would book about 1.7 jobs a month if every one got answered. Take off the jobs you already win without paying anyone, and it can drop under one.

Willison sells round-the-clock coverage, so we have a side in this, and you should read it that way. The method below is built to give the same answer whether the option on the other end is us, a different service, or your own cell phone.

Why is two a week harder to call than twenty?

At twenty night calls a week, the answer is usually clearer, one way or the other. At two, the whole case rests on one or two jobs a month, and a small mistake in any input can flip the answer.

The free option is also a real contender at this volume. Two calls a week is an interruption every three or four nights, and for some owners, taking those on their own cell is the right answer.

So there are really three choices on the table. You answer the calls yourself. You pay a service to take a message, or to patch the real emergencies through to you. Or you pay for coverage that answers, qualifies and books the job. Each one costs something. Only answering it yourself stays off an invoice, and it's paid for in sleep and interruptions instead.

The one line, worked for an HVAC shop

Start with the count. Two night calls a week, weekends included, is about 8.7 a month (two times 52 weeks, divided by 12).

Next, the share of answered calls that turn into booked work. Invoca's 2026 home services benchmarks found that, across the calls it measured, 45% of answered calls to HVAC companies are leads, and 43% of those leads convert on the call. Multiply the two and you get about 19% booked per answered call. That multiplication is ours, not Invoca's.

Put them together: 8.7 calls times 19% is about 1.7 booked jobs a month, if every night call got answered and none of them would've reached you any other way.

Then take off what you already win. Say one night job a month already lands on your calendar, because you called the number back from the log or the caller rang again at seven the next morning. That leaves about 0.7 jobs a month that coverage would add.

Now the other side of the line. Divide the monthly quote by the gross profit on one of those jobs. That's how many added jobs the coverage has to book each month to cover itself. If that number is bigger than what's left after your subtraction, the honest answer for your shop is no.

Where does each number come from, and which way is it usually wrong?

Every input here can be off, and most of the easy mistakes lean toward buying.

The count. Don't count voicemails. Invoca's platform data shows fewer than 3% of callers pushed to voicemail leave a message, so your inbox will badly undercount. Use the carrier or phone-system log instead.

But don't take the raw log either, because it overcounts. Strip out the same number calling three times in ten minutes and anything you know is a robocall. Go easy past that point, though: Invoca's lead share already allows for calls that were never jobs, so scrubbing the log hard and then applying it counts the junk twice. Put your maintenance-agreement members on their own line, since a customer who already pays you for upkeep may be less likely to go looking elsewhere.

The booking rate. Invoca's figures are averages across the calls it measured, answered by a person, and they aren't a night-only number. Night callers might be more urgent than the average caller, or more likely to be shopping around, and nothing published tells you which.

We don't know of a published booking rate for an AI answering at night either, and we don't publish one of our own. So borrowing a rate measured on calls a person answered may flatter any automated option, ours included. Treat 19% as a stand-in, shade it down for an automated answer, and swap in your own daytime rate if you track it.

The value. Use gross profit, not the ticket, and take out what the night itself costs you. A night repair may carry an after-hours fee, but it also carries overtime or on-call pay, and a ticket figure makes every job look bigger than what it leaves you.

What's left out. Some of it leans toward buying: repeat work and referrals from a customer you won at 9pm, and the bad-weather nights the average hides. Some of it leans the other way: callers who'd rather not talk to an automated receptionist, and the time you'll spend telling any service your hours, services and rules, then keeping them current. None of it has a number we can stand behind, so it all stays out of the line. Just know it cuts both ways.

We worked through turning a call log into a count you can trust in how many jobs a two-truck shop loses to unanswered calls.

What you already win comes off the top

This is the step most sales math skips. Coverage doesn't earn every night call. It earns the ones you'd have lost.

Some night callers you already keep. You see the missed call at 8:15, you ring back from the truck, and you book it. Others don't need you to do anything: the homeowner with a noisy blower motor calls again at 7:05 the next morning, because the problem will still be there and they may only have wanted it on somebody's list.

Subtract jobs, not calls. The ones you call back tend to be the promising ones, the numbers you recognize and the voicemails that sounded real, so calling back half your night calls can win you well over half the night jobs.

It also helps to split a booked night call into two kinds. One is work tonight, at whatever after-hours rate you charge. The other is a caller who hears the after-hours option and picks the morning. The second kind may be more likely to have come back on its own, so consider discounting it more.

A plumbing shop at one call a week

Change the trade and the scale. A plumbing shop that gets one night call a week sees about 4.3 a month.

In the same Invoca benchmarks, 30% of answered plumbing calls are leads and 55% of those leads convert on the call. That's about 16.5% booked per answered call, again our multiplication.

So 4.3 calls works out to about 0.7 booked jobs a month, roughly one every six weeks, before you take anything off. If the owner already catches a good share of those by phone, what's left for paid coverage to earn is thin.

That's where answering your own cell often wins outright. The exception is the call that can't wait: water running through a ceiling, a sewer line backing up. One of those in a quarter, landing while you're asleep with the phone on the dresser, can matter more than the average does. Paid coverage isn't the only thing that catches it, either: a service that patches emergencies through, or a forwarded line you'll actually wake up to, can too. Which brings up the average.

Why one month of night calls tells you almost nothing

An average under two booked jobs a month is a small number, and small numbers bounce. Some months will come in at zero and some at three or four, without anything changing in your business. Judge coverage, or the lack of it, on one month and you're mostly reading luck.

Night calls may not arrive evenly either. In the trades that run on weather, they can bunch up in a heat wave or a hard freeze. Two a week across the year can mean a quiet shoulder-season month and a stretch when several come in at once.

That cuts both ways. The average hides the bad night, when coverage is worth the most. But on that same night your trucks are already full. A booked job you can't reach for a day or two is worth less, and some of those callers will move on anyway. Count the bad nights for what your crew could actually get to.

So use several months of log, and the same season a year apart if you have it. A short trial window can't settle this at two calls a week, ours included: a week live shows you two calls, maybe three. That's enough to hear how the calls get handled. It isn't enough to tell you what the coverage is worth.

When the math says don't buy it

Sometimes the honest answer is no. Each of these is a legitimate reason.

• You reliably answer your own phone. If you pick up night calls on your cell and don't resent it, you already have coverage. It's just not on an invoice.
• Your night callers are mostly still there in the morning. If your log shows the same numbers calling back at seven and booking, the subtraction eats most of the gain.
• You don't want night work. If you won't roll a truck after hours and your night callers are mostly emergencies, coverage that books them has nowhere to put them.
• A cheaper service covers what you actually need. If the real emergencies reach you through a message or patch-through service and everything else can wait for a callback, that may be enough. Its quote goes on the other side of the same line.
• The quote needs more jobs than your count can produce. Run the line. If the jobs it has to add are more than what's left after the subtraction, keep your money.

If you already run an on-call rotation, the question changes shape. You're getting woken up for emergencies either way. What paid coverage mostly buys you is not getting woken for the calls that could have waited, and whether that's worth a monthly bill is your call.

For the cheap end of the menu, what sending after-hours calls to voicemail really costs covers why voicemail on its own is the one option where nobody talks to the caller at all.

What Willison does with those two calls

If the line says yes, here's what you'd be paying for. Willison answers every call in seconds, nights and weekends included, asks what's wrong one question at a time, reads the address back, and books the job straight to your calendar. It doesn't tell a caller they're booked until the booking has actually been written.

You decide in daylight what reaches you at night, and it's set per shop. Willison can text you the details so you call back, or, if you turn it on, transfer the live call during the hours you choose. If a transfer rings out, the call doesn't just drop: the details still get taken and sent to you.

It isn't flawless, and neither is any automated receptionist. It can mishear an address, which is why it reads the address back and holds the booking promise until the booking's written. It can misread what kind of job a caller has, which is why it books only the services on your own list and asks the caller to confirm rather than guess, so when your hours or services change, somebody has to update them. And like any system that takes a message, it can fail to get one to you, which is why the owner alert is wired to fire before the caller can leave the call.

The price is flat monthly, with no per-minute meter, no overage and no surcharges, so a busy month or a whole busy season doesn't move the bill. It's scoped to your call volume and needs, and if the business grows into a genuinely bigger operation it gets scoped again, with 30 days' notice and the option to cancel first.

There's a free trial of nine days from checkout: the 48-hour setup, then about seven days live. Payment details are taken at checkout, and the subscription starts automatically when the nine days end unless you cancel. Cancel inside those nine days and you're never charged. It's one trial per business.

If you'd like to hear how a call gets handled before you run your numbers, press play on a real call on willisonhq.com. It's a call to our HVAC receptionist, start to finish.

Frequently asked questions

Get the call log from your carrier or phone system instead. Most carriers and business phone systems can give you a record of calls by date and time, answered or not. The voicemail box will read far too low, since Invoca's platform data puts callers who leave a message at fewer than 3% of those sent to voicemail.

Count them in, but keep them on a separate line. Depending on your hours, "after hours" may take in some or all of the weekend, and a Saturday-morning call is often a different kind of call from a 2am one. Looking at the two groups separately tells you whether you need nights covered, weekends covered, or both.

Often, yes, as long as you actually answer. The trouble spots are the nights you're asleep, driving, on another call or away, when the forwarded call rings out to your personal voicemail. If your log shows you picking up nearly every forwarded call, forwarding may be all you need.

It can tell you how the calls get handled: whether it qualifies properly, reads the address back and books to your calendar. At two calls a week, it can't tell you what coverage is worth, because a week or so live only brings in a handful of calls. Answer the worth question from several months of your own call log.