The call nobody answered is still on your marketing bill

A missed call is a lead you paid for and did not take. Measure it per first-time caller, not per ring: answered, voicemail or missed, and how long until someone reached them. Report each channel's matched revenue beside its answer rate, because a channel whose callers ring after hours will look worse than the demand it creates.

It is 7:40 on a Tuesday evening and a water heater is leaking into somebody's garage. They search, tap the first number that looks local, and listen to it ring six times before a recorded greeting says the office opens at eight. They hang up without leaving a message and tap the second number. That one answers. By the time anybody at the first company sees the missed call the next morning, the job has been quoted, booked and half done by a competitor.

Nothing about that evening shows up as a marketing problem. The ad worked. The click happened, the call happened, the cost was incurred. In the monthly report it appears as a lead that did not convert, charged to whichever channel produced it, and over a quarter it becomes the reason that channel looks expensive. This post is about that gap: what it costs, how it bends channel reporting, and how to measure it from the call log you already have.

What a missed call actually costs

A missed call costs the media that produced it plus the sale it would have become, discounted by the share of missed callers who buy anyway. For businesses that sell by phone, such as home services, clinics, law firms and dealers, it is usually the largest leak between ad spend and revenue.

The media cost is already spent the moment the phone rings, and on some platforms it is charged even if nobody picks up. The sale is the larger number and the harder one to see, because a missed caller rarely complains. They simply call somebody else, and the only trace they leave is a row in a call log marked missed.

Read that the way an owner would. On this channel, a busy receptionist does not just lose the job; the business also pays for the privilege of losing it. The same page says that if the ad schedule differs from business hours, unanswered calls are only charged when they fall inside both, which is a reason to keep those two settings honest rather than leave the ad running into hours nobody covers. Our post on Local Services Ads covers how to reconcile what that channel charges against the jobs it produced.

So on at least one major channel, missed calls cost three times: the charge, the lost job, and a lower position for the next caller. Most channels are less explicit about it, but the second cost applies to all of them.

"Missed" means something different in every phone system

Answered, missed, voicemail and abandoned are not standard terms. Each phone system and call tracker draws the lines in its own place, especially for voicemail and for calls outside business hours, so a missed-call rate means nothing until you know which system produced it and how that system defines it.

None of that is wrong. It is a call-center definition: a missed call is a failure of a person who was on duty, and an abandoned call is a caller nobody was positioned to take. But it means a "missed call rate" pulled from that report leaves out every call that arrived while the office was closed, which for an emergency trade can be the most valuable calls of the week. A marketing question needs the marketing definition: did a human speak to this caller, yes or no.

That is close to everything the measurement needs in one export: who rang, from which source, whether anybody answered, whether they left a message, and whether they were new. If your tracker is CallRail or similar, the fields exist; the work is in deciding what to count. If calls land directly in a phone system such as ServiceTitan's, read its glossary before you trust its percentages.

How missed calls make a good channel look bad

Channels differ in when their callers ring. A channel whose calls cluster in the evening or on weekends loses more of them to an unanswered phone, so it reports fewer sales for the same demand. The distortion lands on the channel, even though the cause is the phone schedule.

Here is the effect worked through. The figures are illustrative assumptions, not benchmarks; rerun the arithmetic with your own. Two channels each cost $9,600 a month and each produce 320 first-time callers. Search ads mostly ring during working hours. Social ads, seen while people scroll at night, ring half the time after hours. The business answers 80% of calls in business hours and 20% after hours, when a single on-call technician is also driving. Answered callers buy 25% of the time, missed callers 5% (some ring back, some get a callback), and the average sale is $800.

The arithmetic. Search ads: 260 × 80% = 208 answered in hours and 60 × 20% = 12 after, so 220 answered and 320 − 220 = 100 missed. Sales are 220 × 25% = 55 plus 100 × 5% = 5, which is 60, worth 60 × $800 = $48,000, or $48,000 ÷ $9,600 = 5.0 per dollar and $9,600 ÷ 60 = $160 a sale. Social ads: 160 × 80% = 128 and 160 × 20% = 32, so 160 answered and 160 missed. Sales are 160 × 25% = 40 plus 160 × 5% = 8, which is 48, worth $38,400: 4.0 per dollar and $200 a sale.

On the standard report, social is 20% worse and is the obvious candidate for a budget cut. Look at the last row. Every answered caller from either channel bought at exactly the same rate. The callers are equally good; the difference is entirely who was around to pick up. Answer every call and both channels produce 320 × 25% = 80 sales, $64,000, about 6.7 per dollar. Cutting social would not fix anything. It would remove the channel that exposed the problem.

The gap can be sized. Each missed caller is worth the difference between the two close rates, 25% − 5% = 20 points. For search that is 100 × 20% = 20 sales, $16,000 a month; for social, 160 × 20% = 32 sales, $25,600. Check: 60 + 20 = 80 and 48 + 32 = 80. Treat those figures as ceilings rather than forecasts. Not every missed caller is a serious buyer: some are spam, wrong numbers or people shopping on price, and the ones who hang up without a message are probably less committed than the ones who wait.

Before cutting a channel for weak return, check its answer rate. A channel whose callers ring after hours can be losing to your phone schedule rather than to its own audience.

Measuring it from a call log

A call log can tell you, per first-time caller, whether anyone answered, whether they left a voicemail, when they rang, which source sent them and how long until someone called back. It cannot tell you why a caller hung up, whether they would have bought, or anything that happened on a phone the log never saw.

The steps below assume an export with one row per call: date and time, the caller's number, the tracking number or source, direction (inbound or outbound) and an answer status. Most trackers and phone systems produce something close to it.

1. Normalize the numbers first. Every comparison below is done on the caller's number, so normalize it to one international format before anything else. An outbound callback dialed from a softphone and an inbound call recorded by a tracker rarely write the same number the same way.
2. Count callers, not calls. One anxious customer who rings four times in ten minutes is one lead, not four missed calls and one answered. Group by number and keep the first inbound call as the event, with the outcome of the whole sequence. The reasoning is the same as for duplicate leads.
3. Separate first-time callers. Existing customers calling about an appointment inflate the call count and are not a marketing loss. Use the tracker's first-call flag where it has one, or exclude numbers already in your customer list.
4. Classify each caller once. Answered on first call; voicemail; missed with no message; abandoned within a few seconds. Write down the threshold for the last one and keep it, because changing it changes the rate.
5. Find the callback. For each missed caller, find the next outbound call to the same number, or the next inbound call that was answered, and record the gap in minutes. No callback is a result too.
6. Mark the hour in local time. Classify business hours and after hours in the location's own time zone. Exports are often in the account's time zone or in UTC, which quietly moves 6 p.m. calls into the afternoon for half your branches.

The third row is where most measurements quietly fail. If technicians return missed calls from their own mobiles, the log shows a missed call and no callback, and the business looks slower than it is. Either route callbacks through the tracked system or accept that callback time is unmeasured, and say so in the report. A number that is wrong in a known direction is still wrong.

Speed to lead: measure from the ring, not from opening time

Speed to lead for phone calls is the time between a caller's first ring and the first conversation with a person. For answered calls it is zero. For missed calls it is the callback gap, and it should be measured from the ring, because the caller was waiting whether or not you were open.

This is the arguable part, and it is deliberate. Many teams measure response time only within business hours, which is how Google calculates the average message response time it shows on Local Services ads when business hours are set. That is fair for a ranking signal and misleading for a revenue question. A caller who rang at 7:40 p.m. and heard back at 8:15 a.m. waited twelve and a half hours. Calling that thirty-five minutes describes your schedule, not their experience.

Report both numbers if you like, labeled, but lead with the one the customer lived through. And report it in buckets rather than as an average: the share of missed callers reached within five minutes, within an hour, the same day, later, and never. An average of two hours can hide a team that calls half its misses back in five minutes and never calls the other half, and those need different fixes.

The value of speed differs by trade, which is why no universal benchmark is quoted here. An emergency plumbing caller rings the next result within minutes, as our post on field services argues; a personal-injury prospect or someone choosing an orthodontist may wait a day and still compare three firms. Match your own callback buckets to your own sales and you will have a better benchmark than any industry survey: the close rate of missed callers reached within an hour against those reached the next day.

Reporting missed calls honestly

Report each channel's revenue as matched, and put its answer rate and missed first-time callers beside it. Show the estimated value of missed calls as its own clearly labeled line. Never add estimated revenue from missed calls into a channel's revenue, because that turns a ceiling into a fact.

It is tempting, once the table above exists, to "correct" social's return to what it would have been with everything answered. Resist it. The $25,600 is an estimate built on two assumed close rates, and blending it into the revenue column makes the report impossible to audit. A finance director can trace a matched sale to an invoice and a phone number. Nobody can trace a sale that did not happen.

• Revenue by channel, matched. Sales whose phone or email appears in the lead file, credited the way you have agreed. Missed callers who bought later belong here, because the channel did produce them.
• Answer rate by channel and by hour. First-time callers answered on first ring, as a share of first-time callers. Split in hours and after hours, because the average of the two hides the story.
• Callback buckets. Share of missed callers reached within five minutes, an hour, the same day, later, never.
• Estimated revenue at risk, labeled as an estimate. Missed first-time callers × the close-rate difference × average sale, with both rates stated.
• Unmatched revenue, stated. Sales with no matching lead stay in their own bucket rather than being spread across channels; our post on unattributed revenue explains why.

CloseRev does not measure answer rates, callback times, call durations or call outcomes, and it has no call tracking or phone-system connection; it reads the files you export. What it does is match: a sale counts toward a channel when its normalized phone or email is in your lead file, credited to the earliest such lead. A missed call still carries the caller's number, so when it is the earliest lead for a buyer, the sale is credited to the source written in that row, and sales with no match stay in Direct / Unknown.

That last point matters more than it sounds. If the export that becomes your lead file contains only answered calls, every missed caller who later bought is either credited to whatever they touched next or lands in the unknown bucket, and the channel that actually produced them loses the credit twice: once for the missed call and again for the sale. Export every inbound call, answered or not.

The missed calls you cannot see

Some missed calls never reach any log you can export: calls to an untracked number, calls from a profile whose platform no longer keeps history, calls returned from personal phones, and calls to a location's direct line. They are part of the gap, and the report should say they are missing.

So the profile, often the largest single source of calls for a local business, no longer tells you which of its calls went unanswered. If the number on the profile is your main line, the only record is your phone system's. Our post on Google Business Profile covers putting a tracking number in its primary phone field while keeping your real number on the listing, which is what makes its missed calls measurable at all.

For a group of branches the blind spots multiply, because each site has its own phones, staff and hours. The same campaign can be answered well in one town and badly in the next, which is why multi-location reporting has to split by site before it judges a channel.

What to fix first

Fix answering before media. The first missed-call report almost always shows a gap after hours or at lunchtime that is worth more than any plausible improvement in campaign efficiency, and it is cheaper to close with staffing, overflow routing or an answering service than with a new channel.

1. Run the measurement on last month: first-time callers, answer rate in and out of hours, callback buckets.
2. Find the hours with the most missed first-time callers. It is usually early evening, Saturday morning or lunchtime.
3. Price the fix against the estimated revenue at risk for those hours, using the ceiling above and then half of it, to see whether the decision survives a skeptic.
4. Make sure missed calls appear in the lead file, so the callers you recover are credited to the channel that sent them.
5. Rerun the report a month later and compare channels again before moving any budget.

If the honest version of the report says your evening channel was fine and your phones were not, that is a good result, not an embarrassing one. It means the marketing was working and the fix is something you control this month. The alternative, cutting the channel whose customers call when it suits them, saves money on paper and makes the business a little harder to reach every time it is repeated.

Missed calls are a marketing number disguised as an operations one. Count them per first-time caller, time the callback from the ring, keep missed calls in the lead file, and report what they might be worth beside the revenue, never inside it.