What a Missed Call Costs You: Do the Sums With Your Own Numbers

Open any page selling call-answering software, human or AI, and near the top you’ll find some version of “you lose X% of callers who don’t reach a live person.” Nobody sources it. The number drifts between 60% and 85% depending on which vendor wrote the page. A figure that wanders that far with no citation under it is a sales device, so we won’t quote one at you.

You’ll get arithmetic instead, with your numbers in it. Ten minutes with your own call log fills the gaps.

Why no honest industry-wide number exists

No credible, sourced figure exists for “the share of missed calls that become lost business”, because the answer swings on things that differ hugely from one shop to the next. Average them together and you get a number describing nobody:

• How many competitors the caller has on hand. Someone with a burst pipe searching “emergency plumber near me” has five more names to try inside thirty seconds. Someone who picked you off a referral, or had you out last spring, will often ring back or wait for your text.
• How urgent the job is. A furnace that died in January gets called back to that hour, by somebody, somewhere. A quote for a kitchen next spring can sit in a caller’s head for weeks before they try again, if they ever do.
• What the caller does next. Some redial straight away. Others text you when your number takes texts, and a fair few drop the whole idea of the job or go back to the search results and work down the list. That behaviour changes by trade, by region, and by the mood of the person holding the phone.

None of that holds still, so no single percentage describes every business reading it. Your own average job value and your own missed-call count sit inside systems you already own. Both are measurable this week.

Step 1: find your real missed-call count

Open your phone’s call log, or your business VoIP system’s call detail report, and count for one full week, two if you can:

• Calls answered
• Calls that went to voicemail
• Calls that rang out with no voicemail left at all

Most owners undercount that last one. A caller who hangs up without leaving a message leaves nothing behind in your day. If your system won’t split answered from unanswered on its own, tally it by hand for a week. The ten minutes pay for themselves here.

Write down the weekly total of unanswered calls. Call this number M (missed calls per week).

Step 2: find your real average job value

Pull your last 20-30 completed jobs from whatever holds your invoices, and average the total billed. Use the real average. Your best month and your flagship job type will both lie to you here. Careful if you run a wide spread, small service calls alongside larger installs: average the specific job type a phone lead is most likely asking about, because a blended average will skew the answer in whichever direction flatters the system you’re weighing up.

Call this number V (average value of a converted lead).

Step 3: decide your own recovery-rate assumption, openly

Every other page on this topic skips this step, because it means admitting nobody knows the answer. You don’t. Neither does the person selling you a system. What you can do is pick a recovery rate you believe, then see what it implies.

“Recovery rate” means the share of the calls you miss today that you would book as a job if you rang back within, say, five minutes, or if a text went out the second the call dropped. Pick a number you’d defend to a business partner who thinks you’re about to waste money. 100% is fantasy, since some callers have already hired someone else by the time your callback lands. 0% would mean a callback never changes anybody’s mind, which doesn’t hold either for most trades. Where you land between the two is a judgement about your own market, and this guide can’t hand it to you.

Call this number R (the fraction of missed calls you believe you’d convert with faster response, expressed as a decimal, e.g. 0.15 for 15%).

Step 4: the arithmetic

With those three numbers, the monthly value of recovering missed calls is:

Monthly recovered value = M × 4.3 × R × V

(4.3 converts a weekly count to an average monthly count.)

Worked example, using round numbers you should replace with your own: a shop misses 8 calls a week (M = 8), the average job is worth $350 (V = $350), and the owner reckons a fast callback or text-back would convert 1 in 5 of the calls that would otherwise be lost (R = 0.20).

Monthly recovered value = 8 × 4.3 × 0.20 × $350 = $2,408/month

Now hold that figure against what a missed-call system costs to run. A narrow, purpose-built missed-call text-back system is a small build, and each message after that costs close to nothing. An AI receptionist that answers the phone rather than texting after the fact runs roughly $79-249/month for a packaged product with unlimited minutes (source: goodcall.com/pricing, checked 2026-08-28), or $125/month plus $0.25/minute overage for a narrower regional product (source: aiphonereceptionist.ca/pricing, checked 2026-08-28). A human answering service handles complicated calls better, and its cost climbs with your call volume: $99-350/month at low volume, into the thousands when the phone is busy (sources: patlive.com/pricing and smith.ai/pricing, checked 2026-08-28).

Against the worked example’s $2,408/month, near enough any of those prices clear the bar. Now picture a shop that misses one call a week on $150 jobs. Very different answer, and that owner needs to run the same sums honestly, on their own numbers.

Step 5: stress-test your own assumption

Run the formula again at half your assumed recovery rate, then at a fifth of it. If it still pencils out at R = 0.05, a 1-in-20 recovery, you’re taking a low-risk decision. If it only works at R = 0.30 or higher, you’re making an optimistic bet, and better to know that going in than to find out six months into paying for a system.

Put your own M and V through that table before you decide anything. The formula sits here in place of a single headline result because your number is the one that counts.

What this arithmetic can’t tell you

It can’t tell you whether your missed calls are simple repeat questions a script can handle or messy conversations that need a person to close. That distinction matters more than the dollar figure, since a system that picks up the phone and then handles the call badly loses you the job as surely as ringing out does. It can’t give you your real recovery rate in advance either, only what a given assumed rate would be worth if it held. You learn the real one by running a system for a decent stretch of time and counting what gets booked, against the missed-call baseline you measured in Step 1.

Run your own M, V and R. You’ll end up deciding on your own numbers, not on a percentage somebody made up for a marketing page.