41% of jobs booked online arrive outside business hours.
If your phone rolls to a generic voicemail at 6pm, you are absent for close to half the demand, and specifically absent for the portion that cannot wait. Nobody with a working furnace fills in a contractor form at 10pm on a Saturday.
The usual response to that statistic is to price a 24-hour answering service, decide it is too expensive, and do nothing. That happens because the statistic hides the important detail: after-hours contacts are not one audience. They are three, and only one of them needs a human right now.
The short answer
Segment the traffic, then buy coverage for each segment separately. Text back and online booking handle the majority at almost no cost. A human answering service is worth paying for only if genuine emergencies are a meaningful share of your after-hours volume and your job values justify it. Buying 24-hour human coverage for a stream that is mostly scheduling is why this looks unaffordable.
The three groups
| Group | What they want | Right response | Cost |
|---|---|---|---|
| Emergency | A human, now, and someone dispatched | Answering service or on-call | Highest |
| Scheduler | To book a time and stop thinking about it | Online booking link | Near zero |
| Researcher | Information, will decide tomorrow | Auto-reply plus a good website | Near zero |
Emergencies are the smallest group and the most valuable. No heat in January, a burst pipe, no power. They will call three companies until someone answers, they will pay a premium, and they are lost within minutes if nobody responds.
Schedulers are usually the largest group and the cheapest to serve. Their boiler is fine; they finally have a moment at 9pm to sort out the service they have been meaning to book. They do not want to talk to anyone. A booking link converts them outright.
Researchers are comparing. They will contact two or three companies and decide in the morning. An immediate acknowledgement holds them; silence loses them to whoever replied first, and around 78% of customers buy from whoever responds first.
By the numbers
Work out your own mix before buying anything. Take a month of after-hours enquiries and sort them into the three columns. Most contractors find emergencies are a smaller share than they assumed, which changes the coverage decision substantially and usually saves money.
The coverage ladder
In cost order. Most businesses should implement the first three regardless and then decide about the fourth.
1. Missed call text back. A small monthly fee, often already included in your phone system or field service platform. Sends an SMS within seconds acknowledging the call and asking what is wrong. This is the highest-return item on the list because roughly 85% of callers will not leave a voicemail and around 85% of missed callers never ring back. Setup and message wording in missed call text back for contractors.
2. Online booking. Free to cheap, and it converts the largest group outright. 80% of homeowners factor online booking into choosing a contractor. Put it on service pages for work that is schedulable without a quote: tune-ups, inspections, standard call-outs.
3. A form auto-reply that states hours and an emergency route. Something like: we have your request, we open at 8am and will reply then; if this is an emergency, call this number. That single sentence sorts your own traffic for you, at no cost.
4. Human answering service. Genuine cost, genuine capability. They answer as your business, qualify the call, and either book it or reach the on-call technician. Worth it when emergencies are a real share of volume and job values are high.
5. AI first response. A middle path that has become viable. 53% of homeowners are comfortable with AI handling an initial enquiry, which is a lower bar than most contractors assume. The homeowner's actual requirement is usually an immediate acknowledgement and a real appointment, not specifically a human voice at 11pm.
6. Full on-call rotation. The most expensive, in money and in staff goodwill. Justified for trades where emergency work is core rather than incidental.
Our free check tests the after-hours side directly: whether your site states hours at all, whether an out-of-hours visitor can book or enquire without phoning, and where that enquiry goes. Across home service sites, 65% show no hours anywhere. Twenty checks, about fifteen seconds.
Capture is not the same as dispatch
The distinction that makes this affordable, and the one most owners conflate.
Capturing an enquiry means it does not go to a competitor. Dispatching means someone drives out tonight.
You can do the first without the second. An enquiry received at 9pm, acknowledged at 9:01pm and booked for 8am is a won job, and the customer is usually perfectly happy, because most after-hours contacts are not emergencies.
Which means "I do not want to work evenings" is a legitimate business decision that does not require going dark. Declining to work nights costs you nothing. Declining to answer costs you the daytime work those people would have booked.
There is a nice illustration of the upside. An operator who built a cleaning business to $60,000 a month described taking bookings between 5pm and 8am specifically because every competitor had stopped answering. His summary of being woken for a job was blunt: it is unpleasant, but it is two hundred dollars.
Most home service jobs are worth considerably more than two hundred dollars.
Set the hours you actually keep
A smaller point that costs real money and takes ten minutes.
65% of home service businesses show no business hours anywhere. A visitor who cannot tell whether you are open at 8pm assumes you are not, and calls someone whose hours are stated.
Worth publishing, in text, on the website and the Google Business Profile:
- Normal hours, accurately. Aspirational hours you do not keep are worse than honest ones.
- Whether you take emergencies, and how. This is a different question from whether you are open, and it is the one urgent callers are actually asking.
- What happens if you contact us out of hours. "Messages received overnight are answered by 8am" is free and it stops people continuing down the list.
- After-hours pricing, at least in principle. Stating a premium exists avoids an awkward conversation at the door and filters people who were never going to pay it.
That last one is worth doing even though it feels like a deterrent. A customer who knows there is a night rate and calls anyway is a customer who will pay it.
Protecting the people who take the calls
The part of this decision that gets discussed least and causes the most turnover.
On-call rotations are genuinely unpopular, and a badly designed one loses you technicians faster than the extra revenue replaces them. If you are going to have people available at night, the arrangement needs to be worth their while and predictable.
Pay for availability, not just for call-outs. Being unable to have a drink, leave the house or fully switch off has a cost even on a night nobody rings. Paying only when the phone goes treats that as free, and people notice.
Rotate predictably and publish it. A rota people can plan around is tolerable; unpredictable availability is not, and it is the single most common reason technicians leave over this.
Filter before the phone rings. This is where the automation pays for itself twice. If text back and booking handle schedulers and researchers, the on-call person is woken only for genuine emergencies. That can be the difference between four disturbed nights a month and one.
Define what counts as an emergency, in writing, and let the person on call decline the rest until morning. Without a definition, every caller's problem is urgent by their own assessment.
Cap the response expectation. "We answer within thirty minutes overnight" is deliverable. "We answer instantly, always" is not, and a promise you break in public is worse than a modest one you keep.
The wider point: after-hours coverage designed around the business alone tends to collapse within a few months. Designed around both the customer and the person carrying the phone, it holds, and the automation layer is what makes the second half possible.
A configuration that works for most small trades
Putting the whole thing together, for a business without the volume to justify full overnight staffing:
Missed call text back, firing within seconds, with an after-hours message variant that offers an urgency escape hatch. Replies land on the phone of whoever is on call.
Online booking live on every service page for schedulable work, taking the largest group off the queue entirely.
Form auto-reply stating hours and the emergency route, so researchers are held and emergencies are redirected.
One on-call phone, rotated weekly, answered for flagged urgent messages only.
Hours and emergency policy published on the site and Google Business Profile.
That combination costs a small monthly fee plus one rota, captures essentially all of the 41%, dispatches only for genuine emergencies, and does not require anyone to be awake and alert every night. It is also considerably cheaper than a 24-hour answering service, which is worth buying later if the emergency arithmetic above justifies it rather than as the first move.
Deciding whether an answering service pays
Simple arithmetic, worth running before you buy or dismiss it.
- Count after-hours enquiries for one month.
- Sort into the three groups.
- Take the emergency count only, since automation handles the rest.
- Multiply by your close rate and average emergency job value.
- Compare against the monthly service cost.
If ten genuine emergencies a month convert at 50% on jobs worth $600, that is $3,000 of recoverable revenue against a service costing a few hundred. Comfortable. If it is two emergencies a month on $250 call-outs, it is not, and text back plus booking is the correct answer.
The full comparison, including where each option fails, is in answering service versus missed call text back.
Measure before and after
Four numbers, taken for a month before you change anything and a month after.
After-hours enquiries received. Split by phone, form and message. This is your baseline and it is usually higher than expected.
Share acknowledged within five minutes. Before coverage, this is typically near zero outside hours.
Share converted to a booked job. The number that justifies the spend.
Emergency share. What proportion genuinely could not wait until morning, which is the figure that decides whether human coverage is worth buying.
Without the before figures you will never know whether the coverage worked, and after-hours spend is exactly the kind of cost that gets cut in a slow quarter because nobody can demonstrate its return. A month of counting protects it.
Where this sits
After-hours is one of six leaks in a typical service business, and it is usually the third largest after missed calls and unfollowed quotes. All six, with formulas to size each in money, are in how many leads is my business losing.
The response-time thresholds that make speed worth this much are in how fast you should respond to a lead, and if your website cannot capture an out-of-hours enquiry in the first place, that is the prior problem, covered in contact form versus phone number on your website.
One closing observation about seasonality, because it changes the priority. After-hours volume is not flat. It spikes exactly when your trade peaks, which for HVAC means the first cold snap and the first heatwave, and October is the busiest month in that trade by operational load rather than July. Coverage that is adequate in April fails in the week you can least afford it, which is an argument for having it running before the season turns rather than reacting during it. The timing logic is in the best time of year to advertise HVAC.