What a missed call after hours costs a small business

The cost has no invoice

A missed call leaves no bill. Nobody sends you a statement for the job that went to the next name on the caller's list. Your phone log shows a number, a time and no conversation, and that looks like nothing.

That is why the cost stays hidden for owner-operators. You see the jobs you did. You never see the jobs that dialed, got nothing and moved on.

Most of the figures online come from companies that sell call answering, so read them with that in mind. The Voksha guide says the average small business loses about $126,000 per year to missed calls, and it credits Invoca and BIA/Kelsey. The Phone2 article repeats the same figure but credits Dialora. Two vendors giving two different origins for one number is a reason to treat it as a sign of scale, not as a measurement of your business. Your own phone log is a better source, and we show how to read it at the end.

After hours is a default, not a bad week

Voksha reports, citing BIA/Kelsey, that 62% of small business calls go unanswered during business hours. After hours it puts the figure at 95 to 100 percent. Read that as a description of how most small businesses are set up. Nobody is assigned to the phone at night, so nobody answers it.

For some trades the evening is when the work arrives. The Phone2 article cites the 2026 Contractor Missed Call Report, which says 35 to 45 percent of HVAC and plumbing calls come in after hours, and that pickup rates in those hours fall below 18%. That is a vendor-republished figure from a single report, so treat it as a pointer. The mechanism is easy to believe, though. A person with water on the floor or no heat does not wait for a business that opens at eight.

The after-hours call is often the most urgent call you get. It is also the one with the least chance of being answered.

Voicemail is not a fallback

Many owners feel safe because they have a voicemail greeting. The callers do not treat it as safety. Voksha says 85% of callers who reach voicemail never leave a message and never call back. Phone2 cites a CallRevu survey, via Nextiva, that puts the figure at 82%. The two numbers differ, but they point the same way: most people hang up.

Speed matters too. Voksha cites an InsideSales.com study of 100,000 sales calls and reports that 78% of customers buy from the first business that responds. The same guide says to call back in under two minutes. A callback the next morning is not a slow version of that. It is a different situation, because by then the caller has usually spoken to someone else.

Voicemail has one real use. It tells you who called. That is worth something, but it is a log, not a recovery.

What sits behind one missed call

The first cost is the job itself. Nextiva gives a restaurant example: five missed calls a day, five days a week, adds up to as much as $1,500 per month in lost sales, on its stated assumption of a 50% close rate. Your numbers will differ, because your job value and close rate are your own.

The second cost is the customer over time. Nextiva describes a hair salon that misses a call and loses a $75 appointment, plus five more appointments a year for an average of five years, plus the friends that client would have sent. For a business with repeat work, the first call is the start of a relationship, not a single sale.

The third cost is the marketing that produced the call. Nextiva puts it plainly: if your online ads run 24/7, so do missed calls. If you pay for ads or local listings that show a phone number at night, you pay for calls nobody picks up.

The fourth cost is yours, and no vendor article counts it. It is the owner who checks the phone at dinner, the callback list that waits at the start of every working day, and the first hour spent on people who have probably already booked elsewhere. We covered the daytime side of this in what answering the phone all day actually costs. The after-hours version is the same labor, moved to the hours when you are meant to be off.

What a system can and cannot own

Part of this can be automated, and part of it should not be. A system can answer, ask what the caller needs, record the details, check a calendar, send a text confirmation and put a short summary in front of you before you start work. It can do that every night without anyone watching it.

It cannot judge every emergency. It cannot quote a job when you have never written down how you price it. It cannot promise that a technician will arrive when no technician is free. Those decisions need a person, and a well-built system hands them to you with the details already written down. We wrote about where that line sits in how to set limits on what an AI agent can do alone.

We build and run systems like this at Precipitate. Today our own operation has 197 scheduled jobs running 24/7 across 78 live integrations. That number describes our own setup, not a result for any customer. Scope and price depend on the project, and the place to start is our contact page. The first step is always to map the manual work and say plainly what a system can own.

Measure your own number this week

You do not need a vendor study. You need your phone log and a pen.

First, pull the last month of call records from your phone provider. Mark every call that arrived outside your working hours. Mark which ones were answered, and which ones led to a voicemail, a callback or nothing.

Second, count how many of the unanswered calls you returned, and how long you took. Calls you never returned are lost for certain. Calls you returned the next morning are at risk, and you can check how many of those turned into work.

Third, take the number of after-hours calls that turned into no work and multiply it by your own average job value and your own close rate. Add what you spend on ads that show your number. If you want the longer view, add your typical repeat work per customer.

Write the final figure on a single line of paper and keep it where you can see it.