A worked calculation from call volume, answer rate, enquiry share and job value, with every assumption exposed, a way to measure your own number in a week, and four fixes priced against the loss.
A 10-person service business taking 25 calls a day loses somewhere between $4,000 and $7,000 a month to calls nobody answered, if its average job is worth $250. That's 17 to 28 jobs a month that rang, got no answer, and went to whoever picked up next. The range depends on one thing: whether the phone is answered live 61% of the time, the figure Invoca found across 60 million calls to staffed businesses, or 38% of the time, the figure a 30-day study of 85 small businesses found. Below is the arithmetic, every assumption exposed so you can swap in your own numbers, and a way to measure yours in a week.
Where the numbers come from
- A 30-day study of 85 small businesses across 58 industries found 37.8% of inbound calls answered live, 37.8% sent to voicemail and 24.3% never picked up. Combined: 62% unanswered.
- Invoca's 2025 benchmark, built on more than 60 million calls, found 61% of callers reaching a person and 37% of phone leads converting during the call. Its home-services report puts lead conversion at 46%.
- Hiya's State of the Call research puts the share of callers who reach voicemail and hang up without leaving a message at about 80%.
- Summaries of 2026 call data place 34% to 40% of business calls outside nine to five. ServiceTitan's home-services data has 47% arriving after 5pm, at weekends or on holidays.
- The Lead Response Management Study (Oldroyd, 15,000 leads) found the odds of qualifying a lead 21 times higher when contact came within five minutes rather than thirty. Harvard Business Review's 2011 follow-up across 1.25 million leads found firms responding within an hour were seven times as likely to qualify the lead as those waiting an hour more, and sixty times as likely as those waiting a day.
Those are the only external numbers in this post. Everything below is arithmetic on top of them, plus assumptions I'll flag.
The worked example: a 10-person business
An illustrative firm, so the sums are visible. Assumptions: 25 inbound calls a working day, 22 working days a month, 40% of calls are new enquiries (the rest are existing customers, suppliers and noise), 37% of answered new enquiries turn into a job, an average job of $250, and 30% of callers who get no answer try again later and get through.
| Staffed office (61% answered) | Typical small business (38% answered) | |
|---|---|---|
| Calls a month | 550 | 550 |
| Calls not answered live | 215 | 341 |
| New enquiries among them (40%) | 86 | 136 |
| Lost for good (70% don't try again) | 60 | 95 |
| Jobs those would have become (37%) | 22 | 35 |
| Revenue at $250 a job | $5,500 a month | $8,750 a month |
| Over a year | $66,000 | $105,000 |
Round the numbers down for the callers who leave a voicemail and do get called back, and the honest range for this firm is $4,000 to $7,000 a month. That's the figure in the first paragraph.
What moves the answer most
Average job value
Same firm, same call pattern, different trade. At $80 a job (a salon, a dog groomer) the loss is $1,300 to $2,200 a month. At $1,200 a job (a plumber quoting boiler work, a clinic selling a treatment course) it's $19,000 to $34,000. The call volume is identical. Only the multiplier changed.
Share of calls that are new enquiries
A business with a big base of existing customers ringing about appointments might be at 20%. A business that advertises and takes most work by phone can be at 60%. Halve or double the table accordingly.
After-hours share
If 34% to 47% of your calls arrive when the office is closed, and nobody answers them, your answer rate is capped below 66% before anyone at a desk picks up the first daytime call. This is why the "typical small business" column exists at all: the office isn't bad at answering, it's closed.
Why a callback doesn't get it back
The instinct is "they left a message, we called back". Two problems. First, four out of five didn't leave one. Second, the ones who did were 21 times less likely to convert by the time you rang, if the ring came half an hour later rather than five minutes. Overnight, the HBR data says the odds fall by a factor of sixty. The caller had their phone in their hand and the problem in front of them. By the next morning they've rung two other firms and one of them answered.
The calls that aren't worth money
Not every missed call is a lost job, and a model that says so is selling something. Hiya's data shows 80% of calls from unrecognised numbers go unanswered mostly because they're spam. Some missed calls are suppliers, some are existing customers who'll ring back, some are wrong numbers. The 40% new-enquiry share and the 30% who try again are there to take that out. If your call log says your real numbers are lower, use them. The method matters more than the defaults.
Measure your own number in a week
- Pull the call log from your phone provider. Any VoIP system, and most mobile plans, will show answered, missed and voicemail by time of day. That gives you the answer rate and the after-hours share directly.
- For one week, tally every answered call as new enquiry or other. That's your enquiry share.
- Count how many of those new enquiries booked. That's your conversion rate; Invoca's 37% is the fallback if you don't want to count.
- Take last quarter's revenue and divide by jobs. That's your average job value.
- Listen to a week of voicemails and count how many callers you never reached. That's your "lost for good" rate.
Then: calls a month × missed rate × enquiry share × lost-for-good rate × conversion rate × job value. Five multiplications. Most owners who do this once stop arguing about whether the phone matters.
Four fixes, cheapest first
| Fix | Cost | What it recovers |
|---|---|---|
| A short voicemail greeting plus an automatic text-back ("Sorry we missed you, reply here and we'll call in 10 minutes") | Free to $20 a month | Some of the 80% who wouldn't leave a message will reply to a text |
| A forwarding rota, so a missed office call rings a mobile | Staff time | Daytime misses. Not the after-hours share, unless someone is on call |
| A human answering service | $250 to $1,725 a month by minutes, $3.30 to $5.40 a minute over the bundle | Every call is answered. Most end as a message, so the callback problem above still applies |
| An AI receptionist connected to your booking system | $25 to $300 a month off the shelf, or $0.10 to $0.25 a minute for a custom build after a one-off fee | Answers every call including after hours, books on the call, takes payment, hands the odd one to a person |
Set the fix against the loss. For the firm in the table, a $99 a month subscription pays for itself on the first recovered job. A custom build at a few thousand pays back in one to two months of recovered jobs at the $250 level, and in weeks at the $1,200 level. The answering service is the one to check carefully: at $720 for 200 minutes it can cost more than it recovers if most of its calls end as messages.
When it isn't worth fixing
Under about 50 calls a month, the arithmetic above produces a small number and a mobile with a good greeting is enough. If your average job is under $50 and your enquiry share is low, the same. And if your missed calls are mostly existing customers who will ring back anyway, the model overstates the loss. Do the week of counting first. The number either justifies the fix or tells you to leave it alone, and either answer is worth a week.
If your count comes out anywhere near the table, see how our AI call answering books and takes payment on the call and book a 30-minute call. We'll run the model on your real call log. And for what each option actually costs to run, the numbers are in what an AI receptionist costs in 2026.
Questions owners ask about missed calls
How much does a missed call cost a small business?
Missed rate times new-enquiry share times the share who never try again times conversion rate times average job value. For a 10-person firm taking 25 calls a day with a $250 average job, that comes to $4,000 to $7,000 a month. At a $1,200 job it's $19,000 to $34,000. Count your own for a week before trusting anyone's number.
What percentage of business calls go unanswered?
A 30-day study of 85 small businesses found 62% unanswered live: 38% picked up, 38% to voicemail, 24% no answer at all. Invoca's benchmark across 60 million calls to staffed businesses found 61% of callers reached a person. Your own phone log will tell you which column you're in.
Do callers leave a voicemail?
Mostly not. Hiya's research puts the share who reach voicemail and hang up without a message at about 80%. Of the rest, the odds of converting them fall sharply with every minute before the callback, which is why a message-and-callback system recovers less than it seems to.
How quickly should a missed call be returned?
Within five minutes if you want the lead. The Lead Response Management Study found the odds of qualifying a lead 21 times higher at five minutes than at thirty. Harvard Business Review's 2011 study found responses within an hour seven times as likely to qualify as an hour later, and sixty times as likely as the next day.
How many calls come in after hours?
Summaries of 2026 call data put 34% to 40% of business calls outside nine to five. Home-services data from ServiceTitan has 47% after 5pm, at weekends or on holidays. If nobody answers those, your answer rate is capped below two thirds before the office opens.
What's the cheapest way to stop missing calls?
An automatic text-back on missed calls costs nothing to $20 a month and recovers some of the people who wouldn't leave a message. Beyond that, a forwarding rota covers daytime, an answering service covers every hour but ends most calls as a message, and an AI receptionist answers, books and takes payment on the call from $25 a month.
