Why your customers won't leave a voicemail - and where they go instead

Fewer than 3% of callers pushed to voicemail leave a message. That figure comes from call-tracking firm Invoca, which wrote in May 2024 that its platform data shows "less than 3% of callers who get pushed to voicemail leave a message". The rest, more than 97%, hang up without one.

My bet is that most of them dial the next plumber on the page, and voicemail makes that quick and painless for them.

How many callers actually leave a voicemail?

Under 3%, on the best-documented figure I could find: Invoca's own platform data, published in May 2024 in an article about home-services businesses. Invoca gives no sample size or method for it, and it sells a tool for chasing missed calls, so read it as a vendor's number.

Voicemail abandonment is the share of callers who hear your greeting and hang up without recording anything. On Invoca's number it runs above 97%.

You will also see 80%, 85% and 86% quoted as the share of callers who hang up on a voicemail greeting. I checked three pages that quote them and got three different stories. One credits its 80% to "Hiya research", with no link. Another credits its 86% to Invoca in 2024 - the same Invoca whose own page puts abandonment above 97%. The third gives its 85% no source at all.

So I read Hiya's report. The 2024 edition of State of the Call surveyed 12,127 consumers in six countries, the UK, France, Germany and Spain among them, and analysed more than 221 billion calls. Its sixteen pages never use the word voicemail. The 2026 edition does mention voicemail, from the other end of the line: 41% of consumers say their default reaction to an unknown number is to let it ring and see whether the caller leaves one. Neither edition counts how many callers leave a message.

If you want a number to plan around, use Invoca's, with its caveats: fewer than three callers in a hundred.

Why won't they leave a message?

Because leaving one costs them effort and buys them an open-ended wait. I found no survey that asks callers why they hang up, so what follows is my reasoning about the caller's position.

• A voicemail is a promise with no delivery date. They cannot tell whether you'll hear it in ten minutes or on Thursday. Dialling the next number settles it in seconds.
• The problem doesn't fit in thirty seconds. "There's water coming from under the boiler and I've turned the stopcock but it's still going" is the start of a conversation, and a recording cannot ask the follow-up question.
• They are already holding a list. If they found you through a search, the same page gave them a handful of other plumbers.
• A beep cannot answer what they rang to ask: can you come today?

Rewording the greeting fixes none of that. A warmer voicemail message is still a voicemail message.

Where does the call go instead?

Probably to the next name on the list, though I could not find a study that tracks it. What can be measured is the size of what walks off.

ServiceTitan's data from more than 3,000 trade businesses that use its software put plumbing's call booking rate at 43% in June 2022, the highest of the trades it reported that month. A booking rate, in ServiceTitan's words, is how many calls a shop turns "into jobs and revenue". North American data, four years old, so a yardstick rather than gospel. If your missed callers would have booked at that rate, roughly two in five of the ones you sent to a beep were a job.

Out-of-hours work is billed higher. Trustoo's Dutch cost guide puts the average plumbing hour at €40 to €80, based on requests through its platform, and lists €80 to €160 for a night or emergency hour without saying where that band comes from. Homedeal's guide puts an emergency hour at €100 to €125 on average. If voicemail answers your phone after hours, those are the calls it is dropping.

Why calling them back so often goes unanswered

An hour later you are an unknown number, and most people do not pick up unknown numbers. Hiya's 2026 survey asked more than 12,000 consumers in six European and North American countries what they do by default when a number they don't recognise rings. Only 14% answer straight away. 28% reject the call, and 41% let it ring to see whether the caller leaves a voicemail; Hiya sums it up as 86% not answering unknown calls. Its 2024 edition tied that to trust: 92% of consumers believe unidentified calls are fraudulent. YouGov found the same reflex in the UK when it polled in December 2021: 32% of Britons never answer an unrecognised number, and another 40% said "not often".

That 41% tells you what to do when you ring back: leave a message. It is what four in ten of Hiya's respondents say they wait for.

Run the sequence end to end. They ring; you're under a sink. They hear the greeting and hang up. You spot the missed call at half past four and ring back from a number they have never saved. Odds are they don't pick up, whatever they think of you personally - and if they do, they may well have booked someone else by then.

For more than 97 callers in a hundred, voicemail records nothing at all. Your missed-call log is all that is left of them.

Is voicemail worse than having no voicemail at all?

In one narrow way, yes: voicemail closes the call. I found no study comparing the two, so take this as argument.

A phone that rings out leaves the caller guessing you're up a ladder, and some may try again. A voicemail greeting tells them, in your own recorded voice, that this business is not available to them right now - a finished answer they can act on by dialling somebody else. It also does something to you. It feels like a safety net, so you stop counting the calls it swallowed.

The honest limit: voicemail still catches someone. The patient caller - the annual boiler service, the bathroom quote for next spring - can afford to leave a message, and my guess is that they make up most of the under-3%. Voicemail catches the callers who could have waited, and loses most of the ones who couldn't.

What actually catches the caller who won't leave a message?

Something that answers. That is the entire requirement, and the options differ mainly in what the caller gets when it does.

Option	What the caller gets	Where it falls down
Voicemail	A beep	Under 3% leave anything at all
Divert to a partner or mate	A human voice	Unless they share your diary, they can't book the job
Missed-call text back	A text, moments later	It fires after the hang-up, when the caller may already be ringing someone else
Human answering service	A real person, on script	Charged per call, so the bill grows with your call volume
AI receptionist	An answer, the details taken, the job booked	A genuinely strange call still needs you

We put the middle two side by side in detail in answering service vs AI receptionist; the bottom row is the one we build. And if the worry is that callers will bolt the moment they realise it isn't a person, that objection has its own numbers: what customers actually do when AI answers.

Do this before you change anything

Open your voicemail box and count the messages from the last month. Then count the missed calls in your phone's call log over the same month, and put the two numbers side by side.

The gap between them is the part of your business you have never seen - and counting your missed calls properly helps you work out what it is worth before you spend anything on fixing it. When the gap is bigger than you'd like, see how the plans work. They start at €69 a month ex VAT, which is less than the bottom of Trustoo's €80 to €160 band for a single emergency hour.