Direct answer: nobody can tell you what a missed call costs your business except your own phone log and your own ticket sizes. An outside estimate may describe a different business, market, or measurement period. The honest method is to count your missed calls for two weeks, price them from the jobs you actually sell, and compare that against whatever you would spend to catch them. This piece shows you how to do that in an afternoon, with a worksheet, and it does not invent a single statistic along the way.
Why is every "cost of missed calls" number you've read suspect?
Search the phrase and you will find confident percentages: how many callers hang up on voicemail, how many never call back, what share of small-business calls go unanswered. Before using a published figure, check who measured it, when, and for which businesses. An accurately reported average can still be a poor input for your own decision.
That matters because the decision you are trying to make — whether to pay for call coverage — is sensitive to the number. If your real missed-call rate is a third of the figure in the article, a service that looked like an obvious win is suddenly a toss-up. If it is double, you are leaving more on the table than you think. Borrowed numbers make you confidently wrong in one direction or the other.
So this article refuses to give you one. Instead it gives you a method whose inputs are all yours.
What do you actually need to count?
Four things, and you already have three of them.
1. Missed calls. Calls that rang your business line and were not answered by a person, during a defined window. Your phone system, carrier portal, or handset log has this. If you run the business from a mobile, the missed-call list is the log.
2. Reached-back calls. Of those missed calls, how many did you (or the caller) successfully connect on later? A voicemail you returned and reached counts. A number you called back and got no answer does not.
3. Your typical ticket. Not your best job — your ordinary one. Pull the last twenty invoices and take the median. If your work splits into obvious tiers (a service call versus an install, a haircut versus a color), use the tier that a first-time caller most often books.
4. Your first-call close rate, roughly. Of the new callers you do reach, what share end up booking? If you have never measured this, estimate it honestly from memory and mark it as an estimate. This is the softest input and the worksheet treats it that way.
Everything else in this article is arithmetic on those four numbers.
How do you run the two-week count?
Pick a normal fortnight. Not the week after a storm, not the week you were on vacation. Then:
- Each evening, write down the number of missed calls on the business line for that day. Most systems show this in seconds. Do not filter out spam yet — note the total and, separately, how many were obvious robocalls or wrong numbers. You want both figures.
- For each real missed call, note whether it was during your stated hours or outside them. This split matters later, because the appropriate fix depends on whether anyone is available then.
- Note whether the caller left a voicemail, and whether you reached them back.
At the end of two weeks you have a small table. Say, hypothetically, it shows 22 missed calls, 6 of them spam, 16 real, 9 during hours and 7 after hours, and you successfully reached 5 of the 16 later. Those are made-up numbers to show the shape of the table. Yours will differ, and that is the entire point.
How do you turn the count into dollars?
Here is the worksheet. Fill every blank with your own figure.
| Line | What it is | Your number |
|---|---|---|
| A | Real missed calls in two weeks (spam removed) | ___ |
| B | Of A, reached back and connected | ___ |
| C | Lost conversations = A − B | ___ |
| D | Your rough first-call close rate for new callers (mark "est." if estimated) | ___ % |
| E | Lost bookings ≈ C × D | ___ |
| F | Median ticket | $___ |
| G | Two-week lost revenue ≈ E × F | $___ |
| H | Monthly ≈ G × 2.17 | $___ |
Worked purely as illustration with the hypothetical table above: A = 16, B = 5, so C = 11. If your close rate for new callers is 40% (est.), E ≈ 4.4 lost bookings. At a $180 median ticket, G ≈ $792 per fortnight, so H ≈ $1,700 a month. Illustrative only — every input is invented to show the arithmetic.
Two cautions the worksheet deliberately builds in. First, line C assumes a caller you did not reach was a lost conversation. Some of them were price-shoppers who called three competitors; some were existing customers who would have called back anyway. The count cannot tell those apart, so treat C as an upper bound, not a floor. Second, line D is where most people fool themselves. If you do not actually know your close rate, write down a range — say 25% to 50% — and run the worksheet at both ends. That range tests your assumptions; it does not prove the true rate lies within it. A decision that only works at the optimistic end needs stronger evidence.
What does the after-hours split tell you?
Go back to your table and look at how many real missed calls fell outside your hours. A customer may call from work, on a break, or after getting home. Your own log will show whether those hours are a meaningful gap for your business. If half your missed calls are after hours, no amount of "answer the phone faster" fixes them, because nobody was there to answer. That is a structural gap, and it points at a structural fix — someone or something that answers when you are closed.
Conversely, if nearly all your misses are during hours, the fix might be simpler and cheaper: a second handset, a receptionist during peak, or a rule that whoever is not on a job answers.
What are you comparing the number against?
Once you have line H, the comparison is straightforward: what would it cost to catch a meaningful share of those calls, and is that less than H?
The options run from free to expensive, and the right one depends on where your misses cluster. A detailed comparison of the options is its own article. For the purpose of this worksheet, the question is only: for the price of the fix, how many of your line-C calls would it have to turn into conversations to pay for itself?
Take our own AI receptionist as the worked example, because we publish the price and can state exactly what it does. It is $200 a month with no setup fee. It answers calls, including after hours, collects the caller's name, contact details, and reason for calling, and books appointments through an approved calendar connection with a text confirmation. At a $180 median ticket and a 40% close rate, roughly three recovered conversations would generate $200 in expected revenue at those assumed inputs: $200 ÷ ($180 × 0.40) ≈ 2.8. That is revenue coverage, not profit breakeven; labor, materials and other job costs still have to be paid. Again, your ticket and close rate replace those figures, and the breakeven moves with them.
Do the same arithmetic for any other option you are considering: divide the monthly cost by (your ticket × your close rate) to estimate the recovered conversations needed to cover the fee in expected revenue. For contribution breakeven, replace the ticket with the amount left after the extra labor, materials and other variable costs of a job. For example, if a $180 job leaves $90 after those costs, a $200 monthly fee and 40% close rate require about 5.6 recovered conversations. Both examples are hypothetical, not promised results. Compare monthly requirements with a monthly version of line C, and distinguish repeated calls from separate opportunities.
What does "answered" actually mean?
One thing we have learned the hard way running phone systems: a call that the carrier marks as "completed" is not the same thing as a conversation. A call can connect, play a greeting, and end with the caller having said nothing — because they hung up, because the audio failed, because the greeting ran too long. If you evaluate any answering option, human or automated, insist on evidence of actual two-way conversation, not connection logs. Ask the provider how they distinguish the two. If they cannot, their "answer rate" is measuring the wrong thing.
Checklist: pricing your own missed calls
- Pick a normal two-week window.
- Log missed calls daily; separate spam from real.
- Split real misses into during-hours and after-hours.
- Record which missed callers you later reached.
- Pull your last 20 invoices; take the median ticket.
- Write down your first-call close rate — as a range if you do not know it.
- Run the worksheet at both ends of the range.
- Treat line C as an upper bound.
- Compute monthly contribution breakeven: monthly fee ÷ (contribution per booked job × close rate).
- Convert your two-week opportunity count to a comparable monthly period, deduplicate repeat callers, and compare. Decide from your data and costs.
Questions people ask
Does a missed call cost the same as a lost job? No. A missed call is a lost chance at a job. The worksheet discounts it by your close rate for exactly that reason.
What if I get a lot of spam? Count it separately and exclude it. If you are paying per call for any service, ask how the vendor treats spam — it can matter to your bill.
Should I include existing customers? Include them in the count, but be aware they are more likely to call back. If you can tag which missed calls were existing customers, run the worksheet on new callers only for a more conservative figure.
Can I skip the two weeks and just estimate? You can, but then you are back to a borrowed number — just borrowed from your own memory instead of an article. Memory over-weights the painful misses and forgets the quiet weeks. Two weeks of counting is cheap.
Where to go from here
If your after-hours misses dominate, compare after-hours call handling options before you buy anything. If you want to see what our receptionist does on a call before trusting the worksheet's breakeven, ask for a walkthrough at (505) 319-5866 or look at the plan details on the pricing page. And if you already have a two-week count, run it through the sheet above and bring the numbers — the conversation is much shorter when both sides are looking at your data instead of someone else's.
Service details and further reading
Offer details checked September 14, 2026: AlgoEthos AI Receptionist and pricing. The worksheet is an illustrative decision aid, not a revenue forecast. Read AI receptionist costs for the existing cost guide.