Ask the owner of a four-truck plumbing company what a missed call costs and you'll get a shrug and a guess. Ask what a truck wrap costs, or a month of Google Ads, and you'll get the number to the dollar.
One of those has an invoice attached. The other gets priced at zero, so nothing ever gets done about it.
First, don't trust anyone's statistic. Including ours
Search for missed-call statistics and you'll find numbers anywhere between 14% and 62%, all delivered with equal confidence.
The spread isn't mysterious once you look at the definitions. CallRail's benchmark data puts home services around 14%, counting a missed call as an inbound call that hits voicemail or gets abandoned. ServiceTitan's industry reporting has cited figures near 60% for contractors whose crews are in the field. Both are measuring something real. They're measuring different things. And both companies sell products that look more necessary when the number is frightening.
Use the published benchmarks for the shape of the problem. Get the number from your own phone system.
Pulling your own number takes about ten minutes
You already have this data sitting in one of three places:
- Your phone system. Most VoIP providers have an unanswered or abandoned calls report in the dashboard nobody opens.
- Your field service software. ServiceTitan, Jobber, and Housecall Pro all have it if call tracking is connected.
- Your mobile carrier log, if the business line still forwards to somebody's cell. Count the missed entries for a month.
Pull one month, then split it two ways: during hours versus after hours, and first-time callers versus existing customers. Those splits change what the right fix is, so don't skip them.
The math
Four numbers, and only the third one is really an estimate.
- Missed calls per month, from the report you just pulled.
- The share you never recover. Some callers try again, some you catch on a callback. Be conservative and assume you eventually reach half of them.
- Your book rate, meaning the share of answered opportunity calls that turn into scheduled jobs. ServiceTitan's published benchmark for residential CSRs runs 65–75%, with strong teams above 85% and weak ones below 50%. If you don't know yours, use 60% and move on.
- Average ticket and gross margin. You know both of these already.
Run it for a company missing 40 calls a month. Forty missed, twenty never recovered, twelve of those would have booked at a 60% book rate. Twelve jobs at a $450 average ticket is $5,400 of revenue, which at a 35% gross margin comes to roughly $1,900 a month, or about $23,000 a year.
Every assumption in that chain is deliberately pessimistic. Halve the recovery assumption, or run it for July when the phone doesn't stop, and it gets a lot less comfortable.
The three fixes, cheapest first
Another person on the phones
Fully loaded, $3,000–$4,500 a month. This is the right answer when your call volume is high during the day and the person will also handle dispatch, follow-up, and collections. It's the wrong answer when the misses are concentrated at 7pm and Saturday morning, because now you're staffing hours you can't fill.
An answering service
Cheap, per-call, live in days. What you get is a message rather than a booking. Somebody still has to call back, and by then the customer has already decided you were closed and rung the next number on the list. Reasonable as a safety net for after-hours overflow, no use against a 30% daytime miss rate.
A voice agent
Answers on the first ring at 2am, asks your intake questions, books into your actual calendar. Typically $800–$1,800 a month for a configured system, ours included. Worth it when the misses are after-hours, seasonal, or during the daytime crush one CSR can't cover alone.
Where a voice agent stops
Any vendor telling you their agent handles everything is describing a demo. Here's the honest ceiling.
It handles the calls that look like each other. What kind of job, where, how urgent, when can we come out. For most trades that's somewhere between two-thirds and three-quarters of inbound volume, and nearly all of the after-hours volume, because people calling at 9pm mostly want to know they've been heard and booked in.
It should hand off immediately, with the context attached, on anything involving a price beyond your published range, a genuine emergency, an unhappy existing customer, or a request that doesn't match a job type it knows. The best feature a phone agent has is knowing when to stop talking and go get you.
What "live in four weeks" actually means
For a business with no IT department, no project manager, and no spare hours:
- Week 1. We listen to twenty of your real calls. Not a workshop, recordings. What people actually ask for is never what the intake form says they ask for.
- Week 2. Build. Your job types, service area, pricing rules, escalation list. Two hours of your time, total.
- Week 3. Test against those recordings plus a set of deliberately awkward ones. It doesn't go live until it handles the awkward ones correctly, which usually means escalating them.
- Week 4. Live on overflow and after-hours only. You keep answering what you were already answering, and widen from there once you've heard it work.
Before any of that, do the arithmetic. If your real number comes back at $400 a month, don't buy anything. That's below the line where a system pays for itself and you'd get more out of fixing your callback discipline. We wrote about finding that line in the two-week audit, and about what keeps a system working after launch in products, not projects.
If it comes back at $2,000 a month, you've been paying that bill every month for years. Nobody has been sending you the invoice.