About half the people who call a home service business never reach a person.
The most quoted numbers on missed calls do not trace back to anything. The math below uses only numbers that do, and ends with what to do about it.
JD Worcester6 min readUpdated
On this page
The cost, worked out
| Callers who reach a person | 52% | Industry average |
|---|---|---|
| Answered calls that are real leads | 38% | Industry average |
| Leads that book on the call | 45% | Industry average |
| Your calls in a month | Your phone log | Unknown |
| Your average job | Your invoices | Unknown |
Calls a month × 48% unanswered × 38% leads × 45% booked × your average job
8 jobs
Why our number is smaller than the ones you’ve seen
If you run a home service business and have been pitched an answering service, you have seen the numbers. 62% of calls go unanswered. 85% of people who miss you never call back. Most callers hang up rather than leave a voicemail. They sit on vendor pages, often next to a calculator that turns them into a big annual loss.
We tried to trace them before using them here, and most of them don’t go anywhere.
The 62% appears to come from a 411 Locals test of 85 small businesses across 58 industries, in which a little over a third of calls reached a live person. That’s a real test. It is also a small one, and it was not about plumbers or roofers. The 85% that never call back we could not find in any study at all. HiGrovi, a contractor software company that compiled the published research on missed calls in August 2026, says outright that the figure has no traceable source. The same company found a “report” of 4.2 million contractor calls circulating under its name. It never published one, and it believes the number was generated by a machine and repeated until it looked like a fact.
So the math above uses one source we could follow to its data. Invoca sells call tracking, and every year it publishes what happens on the calls its customers receive. In the 2026 home services edition, 52% of callers reached a person. Answer rates ranged from 32% to 74% depending on the trade. Of the calls that were answered, 38% were real leads, and 45% of those booked before they hung up.
Two caveats. Invoca’s customers pay for call tracking, which likely makes them more organized than the average shop, so the real answer rate across the industry is probably not better than half. And our math assumes the people nobody answered would have behaved like the people somebody did. Some of them call back. Some of them call the next name on the list. Nobody has measured that split well, so we left it out rather than guess.
Why the first hour matters
The best research on speed is older and was not about phones. In 2011, a team writing in Harvard Business Review audited how fast companies answered web inquiries. Companies that replied within an hour were nearly seven times as likely to reach a decision maker as companies that waited even one hour longer. The average reply took 42 hours.
A phone call is that same inquiry with less patience. The person calling about a leak or a dead furnace is not filling out a form and waiting for a reply. They are scrolling to the next number.
Where the calls actually go
Nobody misses calls on purpose. They miss them on a roof, under a sink, driving between jobs, or at 5:15 when the office has gone home and the phone keeps ringing. In a small shop the person who answers the phone is often the person doing the work. The busiest days miss the most calls, and a calculator never shows that part.
Run it on your own phone first
Before you buy anything, pull the last 30 days of your call log from your phone provider. Count three things: calls you missed, how many of those numbers you called back the same day, and how many called you again on their own.
That’s an afternoon of work, and it turns every industry average above into a number that is yours. If it comes out small, keep your money.
If it comes out big, that is worth a free breakdown. Here is how our teardown machine reads a business before that call.
The fix
Start with the one that costs nothing. Many business phone services can send a text when a call goes unanswered: who you are, that you saw the call, and a question about what they need. A caller who gets that text within a minute knows someone is there. A caller who gets a voicemail greeting does not.
Then cover the hours that matter. Your call log will show when most of the misses land. Put a person on those windows, or an answering service that books into your calendar. One that only takes messages moves the problem from your phone to your inbox.
Then count, once a month: calls missed, missed calls returned within ten minutes, and jobs booked from calls. If the first number falls and the third rises, it is working. If not, you know before you spend more on it.
When not to buy
You are booked out three weeks. More calls are not your problem. Capacity is, and a faster phone only lengthens the wait.
You miss a handful of calls a month and return every one within the hour. Keep doing exactly that. It is working, and nobody should sell you a replacement for it.
Your jobs are small and rarely repeat. Run the formula with your real average job before paying anyone. Eight lost jobs at $90 is a different conversation from eight at $9,000.
Questions owners ask
Don't most people just call back?
Some do. Nobody we could find has measured how many in a way that holds up, which is why the math above does not assume either way. Your own call log will answer it: count the missed numbers that called again within a day.
Is voicemail good enough?
It is better than nothing and worse than a text. Someone who needs help today wants to know they were heard. A greeting does not tell them that. A text back within a minute does.
Should I use an AI receptionist?
Maybe, for after hours and overflow. Test it the way a customer would: call it ten times with ten different problems. If it cannot book the job or hand off cleanly, it is a voicemail with extra steps.
What should I track?
Three numbers, once a month: calls missed, missed calls returned within ten minutes, and jobs booked from calls. That is enough to know whether anything you change is working.
Sources
