The 11pm Booking: Who Calls After You Close

It is 11:04 on a Tuesday night. A customer who has been meaning to book a brake inspection, a haircut, or a table for Saturday finally has a free minute, picks up the phone, and calls the first business that looked right. If you run a Las Vegas auto repair shop, a salon, a med spa, or a restaurant, the question is simple: what does that person get when they call you?

For most small service businesses the honest answer is voicemail, and the published data on voicemail is not kind. This post builds a model you can run on your own numbers: how much demand arrives outside staffed hours, what taking a booking at that moment does to no-show risk, and what it costs to cover those hours with people versus software. Every figure is sourced and dated. Where we assume a number, we say so, and you should replace it with yours.

How much demand arrives after you close

The cleanest figure we have is about calls, not bookings. According to a 2025 Marchex figure summarized on our missed-call statistics page, 34% of calls to service businesses arrive outside standard business hours. We have not seen a comparable published number for self-booked appointments broken out by hour, so we treat calls as the proxy. A phone call at 11pm is a person trying to buy something, and the call log is where that intent shows up first.

The daytime picture is not much better. The 411 Locals study summarized on the same page monitored 85 businesses across 58 industries for 30 days and found 62% of calls went unanswered: 37.8% were answered by a person, 37.8% went to voicemail, and 24.3% got no response at all. 70% of those businesses answered fewer than half their calls. The after-hours gap sits on top of a daytime gap, not instead of one.

What callers do when they hit voicemail is the part that decides the money. About 80% hang up without leaving a message, per the Forbes, Marchex, and BIA figures reported on that page. 75% of callers who cannot reach one business call a competitor instead, usually within minutes (BrightLocal), and 78% of customers buy from the first company that responds (Velocify). Put those three together and the 11pm caller is rarely waiting for your 8am callback. By 11:10 they are somebody else's customer.

Speed compounds it. The Lead Response Management Study (Oldroyd and InsideSales.com, reported in HBR) found leads were 21 times more likely to qualify and 100 times more likely to be reached when the response came within 5 minutes instead of 30. A next-morning callback is roughly nine hours after an 11pm call, far past the 30 minute mark that study already treated as slow.

The after-hours capture model

The missed-call page uses one formula, and we use it here too: monthly calls, times miss rate, times close rate on answered calls, times average ticket, equals revenue at stake. It deliberately refuses the unsourced claim that every missed call costs a fixed dollar amount, and so do we. Your number comes from your call log, not from a headline.

Take a mid-size service shop. Assume 400 inbound calls a month, a 30% close rate on answered calls, and a $250 average ticket. Those three inputs are our assumptions for illustration. If 34% of calls arrive after hours, that is 136 calls a month landing while nobody is at the desk. If the after-hours line is voicemail only, about 80% of those callers hang up without a message: 136 x 0.80 = 108.8 callers who leave no trace except a missed-call entry.

Run the formula on just those silent hang-ups: 108.8 x 0.30 x $250 = $8,160 a month, or $97,920 a year in revenue at stake. If you count all 136 after-hours callers, including the roughly 27 who leave a message but may already have booked elsewhere by morning, the ceiling is 136 x 0.30 x $250 = $10,200 a month, or $122,400 a year. That is gross revenue, not profit, and it assumes those callers would close at your daytime rate, which you should test rather than trust.

The second half of the model is what it costs to answer those hours with people. The hour counts below are plain arithmetic: 2,080 paid hours per full-time employee per year, 7 days x 12 hours x 52 weeks is roughly 4,380 hours, and every hour of the year is 8,760. The $20 an hour is a round placeholder we chose so the math is easy to follow. Our earlier post on the real cost of a front desk works through actual Las Vegas pay; use that figure, or your own payroll, instead.

Going from weekday business hours to round-the-clock coverage adds 6,680 hours, about 3.2 more full-time people, and $133,600 a year in wages at the placeholder rate. That is before payroll taxes, benefits, overtime, and the sick-day coverage a single seat needs. Compare it with the $122,400 ceiling above: in the best case, staffing every hour for the phone roughly matches the gross revenue at stake, and gross revenue is not margin. For this shop, people on the night shift lose money.

Automation changes the denominator. Self-scheduling on your site and profile, or an AI receptionist that can book into your calendar, does not need 8,760 paid hours to cover 8,760 hours. We are not quoting a software price here because none of this post's sources carry one; the point is that the comparison is one tool subscription against three or four salaries, and for most shops that gap is wide enough to decide the question. Our appointment scheduling page shows how we set that up for local businesses.

Where an 11pm booking lands on the no-show curve

Capturing a booking is only half the win. The other half is whether the customer shows up, and the timing of a booking affects that. The best large dataset on lead time we know of is the Resos No-Show Index 2026, which covers 3,768,761 reservations at 2,417 restaurants from August 2025 to July 2026, with an overall recorded no-show rate of 2.33%.

Read the table honestly. A customer calling late at night is most often booking for tomorrow or later in the week, which puts those bookings in the 2.40% and 2.67% bands, slightly above the index average. Capturing late-night demand therefore adds some bookings in the riskiest part of the curve. That is a real cost, and it is small: the difference between 2.67% and the 2.33% average is about a third of a no-show per hundred bookings.

Reminders close most of that gap. In the same Resos dataset, reservations with a delivered SMS reminder showed a 16% lower recorded no-show rate within the same restaurants, a risk ratio of 0.836. Resos labels that an observed association, not proof of cause. As an illustration only, applying 0.836 to the 2.67% band gives about 2.23%, which is under the overall average.

The stronger evidence comes from healthcare. A Cochrane review of mobile phone reminders for appointments pooled 8 randomized trials with 6,615 participants. Attendance was 67.8% with no reminder, 78.6% with an SMS reminder, and 80.3% with a phone call reminder, and SMS versus no reminder had a risk ratio of 1.14 (95% confidence interval 1.03 to 1.26). Clinic attendance rates look nothing like restaurant rates, so do not transplant the percentages, but the direction is consistent: a reminder on a booking made days out is cheap insurance. For what each empty slot actually costs you, see our no-show math for service businesses.

One restaurant-specific note from the same index: dinner reservations recorded a 2.63% no-show rate against 1.84% for lunch. If your late-night bookings skew toward dinner service, expect the higher number and lean on the reminder. We cover how restaurants wire booking and reminders together on our AI for restaurants page.

A second example: the two-chair shop

The first example favored automation strongly because the shop was busy. Small businesses deserve their own run, because the answer can flip. Assume a two-chair salon or a small repair bay with 120 calls a month, a 40% close rate on answered calls, and a $120 average ticket. Again, these are our assumptions for illustration.

34% of 120 is about 41 after-hours calls a month. With voicemail only, about 80% hang up silently: 40.8 x 0.80 = 32.6 callers. Revenue at stake from those hang-ups: 32.6 x 0.40 x $120 = about $1,566 a month, or roughly $18,800 a year. That is meaningful money for a two-chair shop, and it is nowhere near enough to pay a person.

Extending this shop from weekday 9 to 5 to seven days, 12 hours a day, adds 2,300 paid hours: about 1.1 full-time staff and $46,000 a year at the placeholder wage. Even then the phone would still be unanswered at 11pm, because 12-hour days end at night. Hiring cannot solve late-night demand for a business this size. Automation can, but only if it is priced like a small business tool. If a booking system costs more each month than your margin on the recovered bookings, skip it and put the effort into answering daytime calls first, since the 411 Locals numbers suggest that is where most shops leak.

The objection: people who call at 11pm are not serious

The skeptical owner's reply is that late-night callers are browsers, price shoppers, or people who will forget by morning. It is a fair hypothesis, and none of the sources here measure the close rate of after-hours callers specifically. What they do show is that calls in general are high-intent: phone calls convert to revenue 10 to 15 times more than web-form leads, per BIA/Kelsey figures on our missed-call page. Someone who dials instead of filling out a form at 11pm has already chosen the higher-effort channel.

The Resos lead-time table also cuts against the idea that far-ahead or off-hours bookers flake more. The bands differ by fractions of a percentage point, from 1.96% to 2.67%, not by multiples. Whatever else is true of late bookers, they are not showing up at wildly lower rates than everyone else once the booking exists.

The right answer to the objection is not an argument, it is a test. Turn on after-hours self-scheduling for 30 days, tag every booking made outside staffed hours, and compare their show rate and ticket size against daytime bookings. If late bookings genuinely close or show worse in your business, you will see it within a month, and you will have real numbers instead of our assumptions.

Where this analysis does not hold

The 34% figure is a cross-industry number for service businesses. Your own share could be much lower, for example an office that only serves other businesses during the workday, or much higher, for a bar or a late-night food concept. If your call log shows only a handful of after-hours calls a month, the capture model produces a small number and the case for any new tool is weak.

Self-scheduling also does not replace judgment. A personal injury intake, an emergency tow, or a plumbing leak needs a human or a carefully designed routing flow, not a calendar link. For those businesses the after-hours question is about triage and callback speed, and booking software is the wrong tool.

Capacity matters too. If you are already booked two weeks out, extra late-night demand does not add revenue, it pushes bookings further out on the lead-time curve. And be clear about what the evidence covers: the Resos index is restaurants only, the Cochrane review is healthcare only, and neither study measured self-scheduling itself. We use them for lead-time and reminder effects, which is what they measured. The dollar figures in both worked examples are assumptions and should be replaced with your own before you decide anything.

What to do this week

1. Pull the last 30 days of call logs from your phone provider and count calls that arrived outside your posted hours. Divide by total calls and compare your share with the 34% benchmark, then run the formula with your own close rate and ticket.
2. Call your own business at 11pm from a cell phone and listen to exactly what a customer hears. If the greeting does not tell them how to book right now, rewrite it to point to the booking link on your website and Google profile.
3. Turn on SMS reminders for every booking made 2 to 6 days in advance, the riskiest band in the Resos data, and track show rates for a month before and after.

If you would like a second set of eyes on your call log, the free 15-minute audit on our appointment scheduling page walks through it with your numbers.