“I keep missing calls from leads” — what it costs and how to stop it

Missing calls from leads costs you the callers you never get back. Work it out as: missed calls × the share you never reach on a callback × your booking rate × your close rate × average deal value. One person on business hours covers 40 of the week’s 168 hours, so for 76% of the week nobody answers.

• The 168-hour rule: your advertising runs 168 hours a week. A full-time hire covers 40 of them. The other 128 go to voicemail.
• Free triage tonight: ring your own advertised number from a phone that is not yours, pull 30 days of call records, switch on missed-call auto-SMS, re-record the voicemail.
• A callback is not an answered call. Pew Research Center found only 19% of US adults generally answer their mobile when an unknown number calls (n=10,211, July 2020).
• Thresholds: under about 5 missed calls a week is noise and a discipline fix; over about 25 a week is a coverage problem you cannot roster your way out of.
• Three faults hide behind one symptom: coverage, concurrency and screening. Each needs a different fix.

What is actually happening when I keep missing calls from leads

One symptom, three mechanisms, three different repairs. Coverage: the call arrives in an hour nobody is rostered on — evenings, lunch, weekends. Concurrency: the second and third caller arrive while you are on the first, and one handset cannot take two. Screening: you do call back, and your number shows as an unknown caller, so it rings out.

Measure the split before you fix anything. Pull 30 days of call records from your phone provider, VoIP dashboard or call-tracking number and tag every unanswered call as in-hours, out-of-hours or engaged. Those three counts name the gap. Missed calls are rarely one problem, and a roster change will not fix a screening problem.

Is this as urgent as it feels?

Sometimes it is not. Two or three missed calls in a quiet week is noise; the same number every week for six weeks is a trend. Under about five a week, with callbacks going out the same hour, you have a habit to tighten rather than a system to buy. Measure missed calls per week over a rolling six weeks against your own previous six, not against anyone’s benchmark.

Some of this is not a sales problem at all, and ruling it out is free. If your advertised number is not routing — a botched port, a call-forward pointed at a former employee, an expired call-tracking number — no follow-up discipline fixes it and your telco or ad platform is the one to call. If the calls you are dodging are a debt collector, a regulator or a client’s solicitor, the first call is to your accountant or a lawyer, not a marketing supplier.

What to do in the next 24 hours

The first four cost nothing and can be done tonight. Do them before you buy anything.

1. Ring your own number from a phone not on your account — then the number on your ads, your Google Business Profile and your website footer. Confirm each rings a handset a human is holding.
2. Pull 30 days of call records and count misses by hour and weekday. Ten minutes in a spreadsheet tells you whether this is an after-hours problem or a 2pm problem.
3. Turn on missed-call auto-SMS. Most mobile plans, VoIP systems and CRMs have the switch. Plain text works: “Sorry we missed you — this is [name] at [business]. What were you calling about? Happy to sort it here.”
4. Re-record the voicemail greeting so it asks a question and promises a specific time: name, what you need, a callback before 6pm today. Then keep the promise.
5. Set ownership. Write one name against every hour of tomorrow. Ambiguity about who picks up is the most common in-hours cause.
6. Text before you call back. SMS inside five minutes, call attempt inside fifteen. A text is read without the recipient deciding whether to answer an unknown number.
7. Pause ad scheduling on hours you cannot answer. Running click-to-call ads at 9pm with nobody on is paying to generate voicemails.

What is a callback actually worth against an answered call?

Here is the honest answer to “it is fine, we call them back.” A callback is an unlabelled number on somebody’s screen, and most people do not pick those up. In a Pew Research Center survey of 10,211 US adults conducted 13–19 July 2020, 19% said they generally answer their mobile when an unknown number calls, 67% said they do not answer but will check a voicemail if one is left, and 14% said they ignore the voicemail too. A YouGov survey across 17 markets, run online in December 2021, found 26% of consumers never answer calls from unknown numbers and a further 42% do so “not often”.

Be careful with those numbers, because the internet is not: they measure unknown callers in general, not a return call to someone who rang ten minutes ago. A prompt callback is a warmer case. But it is the same unlabelled number on the same screen. Two conclusions follow: text first so your name is attached to the number, and always leave a voicemail, because two thirds of people listen to one even when they will not answer. On how fast, see our lead response time benchmarks.

How much are my missed calls costing me?

Run this on your own numbers, not anyone’s average:

Monthly lost revenue = missed calls × (1 − callback reach rate) × booking rate × close rate × average deal value

Use your own currency; the arithmetic is identical. Treat the result as a ceiling, not a debt: some of those 22 buy later through another channel, and some were never buyers. Halve it and decide on the half. If half still exceeds what covering the phone costs, the argument is over and you are only choosing a method.

The 168-hour rule: why one person cannot cover the phone

This keeps happening because of arithmetic, not motivation. There are 168 hours in a week; a full-time roster covers 40 of them, or 24%.

That 4.2 FTE is the honest cost of solving this with headcount, and it is why most businesses under-roster the phone and then blame the person holding it. Whether software or people should cover the other 128 hours depends on how complex your calls are: we have compared AI voice agents against live answering services on availability, qualification and failure modes.

The next 7 days: the fixes that hold

Triage stops the bleeding. These stop it recurring.

1. Write one answering SLA down. Every in-hours call answered within three rings; every miss gets an SMS inside five minutes and a call attempt inside fifteen. An SLA nobody wrote down is a preference.
2. Route by rule, not by hope. A ring group hitting several handsets at once, then an overflow destination, then the auto-SMS. Sequential hunting to one mobile is where calls die.
3. Log every call against the lead record. If it is not in the CRM it did not happen, and you cannot recount any of this next month.
4. Match ad scheduling to coverage, or coverage to the ads. One of the two, deliberately.
5. Recount at day seven. If misses fell but reach rate did not move, the problem is screening, not coverage, and the fix is follow-up order, not the roster.

On the size of the prize: across the campaigns we run, speed to lead on its own is typically worth about 3x, and doubling contact rate about 2x again. The honest wrinkle is that those do not multiply — 3x by 2x is not 6x, because responding faster is part of how contact rate improves in the first place. Anyone quoting them as a stacked total is selling. The mechanics are in our guide to increasing speed-to-lead conversion rate.

When to fix this by hand, and when to hand the phone over

The honest limitation on the automated option: voice agents are weakest where humans are strongest — distressed callers, heavy background noise, anything needing judgement rather than qualification — so an escalation path to a person is not optional. We wrote that up without the gloss in what AI voice agents can and cannot do. Where the job is to answer, qualify and book, that is what our AI appointment setting work does, paid on booked qualified appointments rather than seats or retainers.

Frequently asked questions

Why do leads not answer when I call them back?

Because your number is unknown to them. In a Pew Research Center survey of 10,211 US adults conducted 13 to 19 July 2020, 19% said they generally answer their mobile when an unknown number calls, 67% said they do not answer but will check a voicemail if one is left, and 14% said they ignore the voicemail too. A YouGov survey across 17 markets, run online in December 2021, found 26% of consumers never answer unknown numbers. Those surveys measure unknown callers in general rather than returned calls, so read them as the upper bound on screening. The practical response is the same: text first so your name is attached to the number, then call, then leave a voicemail.

How many missed calls a week is normal?

There is no credible published benchmark, and any vendor quoting one is guessing. Benchmark against yourself: count missed calls per week for six weeks and compare with your previous six. Under about five a week is usually noise. Over about 25 a week is a coverage problem a roster will not solve, because a full-time person covers 40 of the week’s 168 hours.

Should I text or call back a missed lead first?

Text first, inside five minutes, with a question rather than a statement. Then attempt the call inside fifteen minutes and leave a voicemail if it rings out. A text is read without the recipient deciding whether to answer a number they do not recognise, and it attaches your name to the number before your callback appears.

Do I need to answer calls after hours?

Only if you advertise after hours, or your buyers genuinely ring then. Check the hour-of-day split in your call records first. If after-hours volume is small, the free fix is to match ad scheduling to the hours you can answer. If it is large, that is 128 hours a week you are paying to send to voicemail.

Will an AI voice agent stop me missing calls from leads?

It removes two of the three failures: coverage, because it answers at 3am, and concurrency, because it takes ten simultaneous callers on the first ring. It does not remove the screening problem on outbound callbacks, and it is weakest on distressed or ambiguous calls, so a human escalation route is required, not optional.