Small Business AI / After-Hours Calls

After-Hours Calls: Where Small Businesses Lose Jobs They Paid to Generate

Roughly 40% of calls arrive outside 9-to-5. See what those missed calls cost, why voicemail rarely recovers them, and the honest options for after-hours coverage.

By Adam Hall

TL;DR: Roughly 40% of the calls your business gets arrive outside 9-to-5, and for home services the share runs closer to half. Those callers bought from you already in one sense: you paid for the ad, the listing, or the ranking that made them call. When nobody answers, 85% never call back and most dial your competitor within minutes. Voicemail captures a fraction of it. The fix is coverage, not more ad spend. One disclosure before the numbers: I build AI receptionist systems, so I profit when you conclude coverage matters. Every stat below comes from third parties, and I flag where the sourcing is thin. Discount me accordingly.

About 40% of the calls a small business receives arrive outside standard 9-to-5 hours, and nobody is there to answer them. Of those callers, about 85% will never call back, and most contact a competitor within minutes. If the call came from a paid channel, you spent $35 to $95 to make that phone ring and got nothing for it. After-hours calls are where the marketing budget you already spent goes to die.

That is the short version. Here is the longer one, with numbers you can check.

When your phone actually rings

Business owners think of their call volume as a 9-to-5 phenomenon because that is when they watch it. The phone log says otherwise.

CallRail's analysis of small-business call timing breaks a typical week down like this:

Add the rows outside 9-to-5 and you get roughly 40% of all inbound calls landing when a standard front desk is dark. The single biggest after-hours block is not the middle of the night. It is 5 to 10 PM on weekdays, when your customers get off work and finally handle their own to-do lists.

That 40% is an average, and averages hide the businesses this article is really about:

BrightLocal's study of 45,000 local business listings found restaurants take 51% of their calls after 5 PM, and locksmiths take 34% after 5 PM plus another 8% before 9 AM, with about a third of their weekly volume on weekends. If you run an emergency trade and only answer during business hours, you are dark for the majority of your inbound demand.

Stat callout: ~25% of all calls arrive on weekday evenings alone. The most expensive voicemail in your business is the one that picks up at 5:01 PM.

You already paid for that call

Here is the part most missed-call math skips. Calls are not free to generate.

If you run Google Local Service Ads, you pay per contact. Not per click, per call or message. A February 2026 benchmark from SearchLight Digital, tracking $6.72 million in spend across 888 contractors, put the average LSA lead at $53. By trade: HVAC runs $45 to $85, plumbing $40 to $75, roofing $50 to $95. If you run standard Google Ads instead, the blended cost per lead in home services is closer to $104, and $149 for non-branded campaigns.

So walk through what actually happens at 7:42 PM on a Tuesday:

1. A homeowner's water heater starts leaking.
2. They search, see your Google Verified listing at the top, and tap call.
3. Your phone rings. Nobody answers. Voicemail.
4. Google charges you ~$53 for the lead anyway. LSA bills per contact, and a missed call is still a contact.
5. The homeowner calls the next listing. They answer. Job gone.

You paid for the privilege of losing that job to a competitor.

It gets worse. Google's LSA ranking algorithm uses how fast you respond to leads and your listed business hours as ranking inputs. Google also analyzes calls to detect whether you actually engaged the customer. Miss enough evening calls and you are not just losing the jobs you paid for. You are paying again in worse placement, which raises what you pay per lead next month.

This is why I call after-hours the double-pay problem: once for the lead you lost, once in the ranking that made leads cheaper.

Stat callout: Average LSA lead: $53. Average response-time factor in LSA ranking: how fast you answer. Missed calls cost you the lead fee and the ranking.

After-hours callers are your best callers

Losing 40% of calls would be bad enough if they were average calls. They are not.

Urgency concentrates after hours. The person calling a locksmith at 11 PM or a plumber on Sunday morning is not comparison shopping. They have a problem that cannot wait, they will pay a premium to solve it, and they will hire whoever answers first.

The behavioral data on what those callers do next is consistent across sources:

• 85% of callers whose call goes unanswered never call back (PATLive)
• ~80% of callers sent to voicemail hang up without leaving a message
• 62% of callers who can't reach you contact a competitor immediately (Dialzara)
• ~78% of customers buy from the first business that responds

And response speed compounds all of it. The MIT/InsideSales Lead Response Management study (2007, the source of the famous numbers, often miscredited to Harvard) found that the odds of contacting a lead drop 100x and the odds of qualifying drop 21x when you respond in 30 minutes instead of 5. That study measured web leads, but the principle is the same one your phone log already shows: a caller's patience is measured in minutes, and an after-hours caller's patience is measured in rings.

Harvard Business Review's separate 2011 audit of 2,241 companies found the average business takes 42 hours to respond to a lead, and 23% never respond at all. Your competitors are slow. That is your opening, if you are there at 7:42 PM.

Put it together and the after-hours call is the highest-intent, highest-value, least-forgiving call your business receives. Voicemail is not coverage for that call. Voicemail is where that call goes to convert for someone else.

The math: what the dark hours cost

The formula is the same one from my missed-call cost article, with the after-hours filter applied:

After-hours calls per month × share you fail to convert × booking rate × average job value = monthly revenue lost

Worked example, with assumptions stated so you can swap in your own:

• 250 inbound calls per month (a modest service business)
• 40% arrive after hours: 100 calls
• Voicemail and next-day callbacks realistically convert 15–20% of them, so call it 80 effectively lost
• Booking rate on answered calls: 35% (conservative for inbound)
• Average job value: $450

80 × 0.35 × $450 = $12,600 per month. About $151,000 per year.

That number deserves suspicion, and you should apply it. Not every one of those 80 callers was bookable. Some were existing customers, vendors, spam. Discount it hard. Take a third of it: still ~$50,000 a year, sitting in the hours when nobody picks up. That conservative floor is consistent with the $30,000 to $80,000 annual range the broader missed-call data supports.

Now layer the ad waste on top. If 20 of those 100 after-hours calls came through LSA at $53 each, that is $1,060 a month spent on leads your voicemail processed. Call it ~$12,700 a year in media spend that produced nothing, before counting the ranking penalty.

Same assumptions throughout: 35% booking rate, $450 average job. Your numbers will differ. The point is the shape of the problem, and the shape does not change much when you argue with the inputs.

Why you can't staff your way out of this

The obvious fix is answering the phone. The economics say otherwise.

Covering 24/7 with humans means three shifts. At $50,000 to $70,000 loaded cost per CSR, that is $200,000 or more per year, before training, before turnover, and before you solve the separate problem of finding people willing to answer phones at 2 AM for $20 an hour. That pencils for a call center. It does not pencil for a 5-person plumbing company.

The standard workarounds each break in their own way:

• Forward to the owner's cell. Works until it doesn't: dinner, sleep, the job site. It also trains your family to resent the phone and trains you to answer with the wrong tone at 11 PM.
• On-call rotation across techs. Reasonable for true emergencies, but it burns out your best people on calls that turn out to be routine, and it slows the job they are actually on when the phone rings.
• Voicemail with a fast callback promise. The callback finds a customer who already booked someone else. Callback conversion from voicemail runs around 11–20%.
• Hire a night-shift answering service. Real humans, but they take messages. Message-taking is not booking, and the next-day callback hits the same 85% wall.

None of these are moral failures. They are cost structures, and each one leaks somewhere specific. The question is which leak you can afford.

The four real options, honestly priced

I covered the full cost comparison in the AI receptionist vs. answering service article, so here is the after-hours-specific version:

The last row is what I sell, so here is the disclosure again: $1,500 setup, from $200 a month, with usage billed at cost (roughly $0.10–$0.30 per minute, configuration-dependent; high-volume shops should be on a pooled or unlimited voice plan, and I will tell you which based on your actual call log, not my margin). For a business with 100 after-hours calls a month, the all-in number typically lands between $300 and $600. Compare that to the conservative $4,200 loss row in the table above and the decision makes itself. Compare it to nothing, if your call log shows after-hours volume is trivial, and the honest answer is to spend nothing.

That is the actual test. Pull your call log for the last 90 days. Count what arrived after 5 PM and on weekends. If the number is small, close this tab. If it is a third or more of your volume, you are paying for leads you then refuse to collect.

What this is NOT

• Not a claim that every after-hours call is gold. Some are spam, some are existing customers with questions a business-hours callback handles fine. The math above already discounts for this; discount it more if your mix warrants.
• Not a promise that coverage fixes a broken business. If your close rate on answered calls is bad, answering more calls loses money faster. Fix the sales problem first.
• Not an argument that AI is the only answer. For low-volume shops, a disciplined same-evening callback habit beats a badly configured bot. For true emergency trades, a human on-call remains part of the design. Anyone who tells you one tool covers every case is selling you something. Including me, which is why the call-log test comes first.
• Not a substitute for reading your own data. The 40% figure is an industry average. Your call log is your truth.

FAQ

What percentage of business calls come in after hours?

About 40% on average, per CallRail's small-business call timing analysis: ~25% on weekday evenings, ~8% overnight, ~7% on weekends. The share varies widely by industry. Home services run near 47%, and emergency trades like HVAC and plumbing see 55–70% outside business hours.

Do after-hours callers leave voicemail?

Mostly no. Around 80% of callers who reach voicemail hang up without leaving a message, and 85% of callers whose call goes unanswered never call back. An after-hours caller with an urgent problem moves to the next listing within minutes, so voicemail captures only a small fraction of that demand.

Are after-hours calls really worth more than daytime calls?

Generally yes, because urgency concentrates after hours. A caller at 9 PM usually has a problem that cannot wait, is not price shopping, and hires whoever answers first. That is why roughly 78% of customers buy from the first business that responds, and why missing these calls hurts more than the raw volume suggests.

Is an answering service enough for after-hours coverage?

It beats voicemail, but understand what you are buying. Answering services take messages; they rarely book jobs. Since most callers contact a competitor immediately and buy from the first responder, a message that produces a next-day callback still loses most of the value. Coverage that books is worth more than coverage that transcribes.

Does missing after-hours calls hurt my Google ranking?

For Local Service Ads, yes. Google uses your response speed and listed business hours as ranking inputs, and analyzes calls to check whether you engaged the customer. Consistently missing calls degrades placement, which raises your cost per lead. You lose the job you paid for, then pay more for the next one.

What is the cheapest way to cover nights and weekends?

Cheapest is a same-evening callback discipline on voicemail, which is free but converts only 15–20%. Answering services run $250–$875 a month for message-taking. A managed AI receptionist that actually books runs from $200 a month plus usage. Pull your call log first; the right spend depends on your volume.

The bottom line

You do not have a lead generation problem. If the phone rings at 7:42 PM, marketing worked. You have a lead collection problem, and it costs you twice: the job you already paid to generate, and the ranking that would have made the next one cheaper.

The fix starts with a free step: pull 90 days of call logs and count what arrived after hours. That number tells you whether this article describes your business or somebody else's.

If the number is ugly, I am happy to look at it with you. No pitch deck, no pressure. We count the calls, run your math with your job values, and if coverage does not pencil, I will tell you that too.

Read next: How Much Does a Missed Call Cost a Small Business? | AI Receptionist vs. Answering Service | AI Receptionist for Plumbers | Contact Adam

Sources

• CallRail small-business call timing analysis (window breakdown: ~60% business hours, ~25% evenings, ~8% overnight, ~7% weekends). Provenance note: figures circulated via secondary aggregations of CallRail data; directionally consistent across sources, treat exact percentages as approximate.
• BrightLocal, study of 45,000 local business listings (restaurants 51% of calls after 5 PM and 32% on weekends; locksmiths 34% after 5 PM, 8% before 9 AM, 31% weekends). First-party study.
• SearchLight Digital LSA benchmark, February 2026 ($6.72M spend, 888 contractors, $53 average cost per lead; trade ranges via HomeServiceDirect.net 2026 guide). Provenance note: reported via secondary coverage; methodology summary available, full dataset not public.
• LocaliQ analysis of 3,211 home services ad campaigns, April 2024–March 2025 ($104 blended CPL, $149 non-branded; CPC figures by trade).
• MIT/InsideSales.com Lead Response Management Study, 2007, Dr. James Oldroyd (100x contact odds, 21x qualification odds, 5 vs. 30 minutes). Primary source. Frequently miscredited to Harvard; HBR's 2011 study is the separate 42-hour average and 7x-within-an-hour findings.
• Harvard Business Review, 2011 audit of 2,241 companies (42-hour average response, 23% never respond).
• PATLive (85% of unanswered callers never call back); Forbes/industry data (~80% hang up on voicemail); Dialzara (62% contact a competitor immediately); first-responder purchase rate ~78%. Provenance note: widely cited industry figures; consistent across sources, treat as directional.
• Industry staffing cost estimates ($50–70K loaded CSR cost; ~$200K+/yr for 24/7 three-shift coverage).

Bias disclosure: Adam Does AI builds and manages AI receptionist systems. Pricing claims about my own services are current as of publication and stated openly; all third-party statistics are cited above so you can check them without trusting me.