Here is a Monday I have watched happen at more than one company. The phone system says 25 missed calls. Somebody pulls the list at lunch, splits it between two reps, and by the end of the day three of those people have been reached. Not three sales. Three conversations. The other 22 either went to voicemail, called a competitor, or decided the ants could wait a week.
The part that bugs me is that 25 was never the real number. It was the number the phone system happened to count. The real number is higher, and until you know it you can't know what missed calls are costing you or whether the fix is working.
This post covers how to get the true miss rate, what a missed new-customer call costs, why speed matters more than the callback itself, and the cheapest fix I know of: an automatic text that goes out within a minute of the miss.
Why the phone system undercounts
Every phone system has a "missed calls" report. Every one I have looked at closely leaves things out, because "missed" means something narrow to the vendor and something broad to you. To you a missed call is any person who wanted to talk to your company and didn't. To the phone system it is usually a call that rang an agent and timed out. Those are different sets.
The gaps I see most often:
- After-hours calls. If your system routes to a voicemail greeting at 6:01pm, that call is "answered" by the greeting. It shows up as a voicemail, or nowhere. It is a missed call.
- Abandoned in queue. The caller heard hold music for 45 seconds and hung up. Many systems file that under abandonment, which lives on a different report from missed calls, if it lives anywhere.
- Forwarded to a cell and dropped. Small offices forward the overflow line to the owner's phone. The phone system logged a completed forward. The owner was on a ladder and never picked up.
- Answered by the wrong thing. An IVR that took three presses and the caller gave up. An auto attendant that sent them to a mailbox that nobody checks. Technically connected. Practically missed.
- Answered and then not really. A rep picked up, put the caller on hold to check the schedule, and the caller hung up. Answered, in the report. Lost, in reality.
Add those up and I would guess most offices miss 30 to 50 percent more calls than their phone report admits. That's a guess, not a study. The next section is how you replace the guess with your own number.
How to measure the true miss rate
You need two sources: the raw call log and the recordings. The call log gives you every inbound attempt with a timestamp and a duration. The recordings tell you what happened on the ones that connected.
Step one: pull every inbound call for one week
Not the missed-call report. The full inbound log. Export it to a spreadsheet with the caller number, the time, the ring duration, the talk duration, and the final disposition (answered, voicemail, abandoned, forwarded). If your system won't export that, that is a finding on its own.
Step two: classify anything under 30 seconds of talk time
A real new-customer conversation is rarely under 30 seconds. Filter the log to every call where talk time is short or zero, and sort by disposition. You'll find the after-hours voicemails, the queue abandons, and the forwards that never picked up. Count them. That is your first correction.
Step three: listen to a sample of the short answered calls
Take 20 or 30 calls that were answered but short. Some are wrong numbers and vendors. Some are a customer who got put on hold and left. Some are a rep saying "can I call you right back" and never doing it. Each of those last two is a missed call in every way that matters, and only a recording will show you which is which.
Step four: separate new customers from existing ones
Match the caller numbers against FieldRoutes. An existing customer who missed you about a reschedule matters, but it is a service problem, not a sales problem. The number you care about most is missed calls from phone numbers you have never seen. Those are the ones you paid marketing money to generate.
When you're done you have four numbers for the week: total inbound, true misses, true misses from new numbers, and the share of those that ever got a callback. Most companies that do this for the first time find the last number is under 20 percent. That is the Monday from the top of the post, measured.
What a missed new-customer call costs
Operators I talk to usually put their cost per lead somewhere between $75 and $150, depending on market and channel. Lead aggregators sit at the top of that range and referrals below it. Use your own number if you have it. If you don't, pick $100 and adjust later.
A missed new-customer call is a lead you already paid for and did not get to pitch. If you miss 15 of those a week and reach back to two, you wrote off 13 leads. At $100 each that is $1,300 a week in marketing spend that produced a ring and nothing else. Call it $65,000 a year, before you count the revenue the ones who would have bought were worth. In pest control that is a recurring contract, not a single job, so the lifetime number is several times the lead cost.
I would treat those figures as illustrative. Your miss rate, your lead cost and your close rate all move the answer. The point is that this is not a rounding error, and it is not a customer service problem. It is a marketing problem with a phone-shaped hole in it. I wrote about the same math for leads that connect and then go quiet in the cost of missed follow-ups. Missed calls are the version where the lead never even got a first touch.
Speed to lead: minutes, not hours
The instinct is to fix missed calls with a callback list. Somebody works it after lunch. The problem is that the odds of reaching the person, and the odds of them still being a lead when you do, collapse within minutes of the miss.
The research on this is old but consistent. The Lead Response Management study, done with James Oldroyd (then at MIT) on three years of data across six companies, more than 15,000 leads and over 100,000 call attempts, found that the odds of qualifying a lead dropped 21-fold when response time went from 5 minutes to 30 minutes. Most of the drop happened between 5 and 10 minutes. Oldroyd, Kristina McElheran and David Elkington later wrote up the same theme for Harvard Business Review in 2011 under the title "The Short Life of Online Sales Leads": most companies do not respond nearly fast enough.
Those studies were about web leads, not pest control phone calls, and the numbers are from a different decade. I still believe the shape. A homeowner with a wasp nest on the porch is calling the next company on the list about four minutes after your voicemail picks up. If you're still "working the list after lunch," the list is mostly people who already booked someone else.
The text-back fix
The cheapest thing that works is an automatic text message sent to the caller within a minute of the missed call. It does not need to be clever. It needs to do two things: prove a human-run company noticed, and give the person a reason to reply instead of dialing the next number.
A version that works:
Hi, this is Dana with Ridgeline Pest. Sorry we missed your call just now. What are you seeing at the house? I can usually get someone out today or tomorrow.
Pull that apart. It is from a named person, so it reads like a text and not a notification. It apologizes once, briefly. It asks a single question the caller can answer with one thumb: "ants in the kitchen" or "I think we have mice." And it names a time, which is what the caller was calling to find out anyway.
What not to say:
- Don't send a link. A link in a first text from an unknown number looks like spam and gets ignored or reported. The reply is the goal, not a click.
- Don't ask two questions. "What's the issue and what's your address?" gets fewer answers than either alone.
- Don't quote a price. The caller hasn't told you what the problem is. A number now anchors the conversation before you know what you're selling, and if a system quotes it you may be stuck honoring it.
- Don't say "someone will be with you shortly." That is the hold-music message in text form. It promises nothing and asks nothing.
- Don't send it to existing customers with the new-customer script. Match the number against FieldRoutes first. An existing customer gets "Sorry we missed you, is this about your service on Thursday?"
Timing matters as much as wording. Within a minute is the target. Within five is fine. Fifteen minutes later and you're competing with whoever they called next.
Routing replies to a human
The text-back is the easy half. The hard half is what happens when the person replies "ants in the kitchen, how much?" and nobody is watching the inbox. An automatic first message with a manual second message is the same missed call with extra steps.
You have three options, and which one you pick depends on your office.
- A shared inbox one person owns during business hours. This works for offices with a CSR who is at a desk all day. It does not work after 5pm, which is when a lot of the misses happen.
- Round-robin to whichever rep is free, with an alert in Slack or on their phone. Better coverage, and it makes the reply visible to a manager. The failure mode is that everyone assumes someone else has it.
- An AI text agent that handles the conversation to a booked appointment and hands off when it gets stuck. This is what covers nights and weekends. The handoff rules matter more than the AI: it should pass to a person on any price dispute, any complaint about past service, anything about pets or chemicals it isn't sure about, and anyone who asks for a human.
Whatever you choose, close the loop in FieldRoutes. The reply should create or update the lead, and the booked appointment should land on the right route. If the conversation lives only in a texting tool, you have built one more place to look and one more report that disagrees with the others.
Compliance, briefly
Texting back someone who just called you is the friendliest case in messaging compliance. They initiated the contact. Replying by text to an inbound inquiry, about the thing they inquired about, is fine. Two things still apply. Honor opt-outs immediately: if someone replies "stop," they get nothing further, and your system should enforce that rather than relying on a rep to remember. And keep the text about their inquiry. The person who called about termites at 7pm did not sign up for your spring mosquito promotion. Existing customers on an opted-in list are a different situation with its own rules, and I'm not a lawyer, so if you are unsure about your state, ask one.
How Plaibook does it with FieldRoutes
Since people will ask: this is one of the things we built Plaibook to do, and the measurement side came first. Plaibook pulls the call log and recordings from your phone system (Five9, Dialpad, RingCentral, CallRail, CallTrackingMetrics, Quo, Nextiva and a dozen others) and matches every inbound number against FieldRoutes. The missed-call tracker counts the misses the phone report skips: after-hours, abandoned, dropped forwards, and the "answered" calls where the caller left before anyone helped them. It splits them into new numbers and existing customers, shows which rep or queue missed them, and shows what share got a callback and how long it took.
The text-back runs on the same data. When a new number misses you, Plaibook sends the first text within a minute, from a name you choose, with a question and a time offer written in your words. When the person replies, the AI carries the conversation, handles the common objections, and books the appointment on the right route in FieldRoutes. Anything it shouldn't handle goes to a person. If you want a human to approve every message before it sends, co-pilot mode does that, and a lot of teams start there. There is no manual entry for the rep in any of it, and FieldRoutes stays the source of truth.
You can read how the texting side works on the AI text follow-up page. But even if you never use us, do the one-week audit. Pull the full log, count the real misses, listen to the short calls. The number is almost always worse than the phone system says, and you can't fix a Monday you haven't measured.