Your Missed Calls Are Closed-Lost Jobs — Start Tracking Them

My NPS in month two of running Reeves Electric was a 4. Not 40. A 4 — on a scale where anything below 50 means your customers are actively telling people to avoid you. I didn't get there by doing bad electrical work. I got there because customers were disappearing before they ever made it into my CRM, and I had no idea.

Here's what I thought was happening: people called, occasionally missed us, left a voicemail, we called back, we booked the job. That was my mental model. It was completely wrong.

When I started recording every call in those first ninety days, I pulled the missed call data and looked at how many produced a voicemail. Almost none. Not "a few." Almost zero. Dead line, end of story. The customer didn't complain. No one-star review. They just called someone else, and I never counted the loss because the job never made it into my system.

---

What Happens When You Miss a Call

The miss doesn't feel like revenue leaving. That's the problem. When you lose a job you quoted, it shows up somewhere — the estimate that didn't convert, the customer who went with the other guy. You know. When you miss the call before the quote, nothing shows up. The job never existed in your system.

What actually happens on the other end is simple. Someone searching "electrician near me" clicks your LSA listing, hears four rings and no answer, and dials the next result. That shop picks up. Books Thursday. Sends a confirmation text. You don't know any of this happened because nothing in your data changed.

The month-two NPS of 4 forced me to stop assuming I knew what was going on and start listening. What I heard, once I had the recordings, was calls landing between 1pm and 4pm on days I was under a panel — and just ending. No voicemails. No callbacks.

---

The Number You're Not Running

Here's the math I did when I finally ran this. My average ticket at the time was around $680. My booking rate on calls I actually answered was around 55%. I was missing, conservatively, 12 calls a week.

Twelve missed calls × 55% booking rate × $680 average ticket × 52 weeks. Roughly $233,000 in annual revenue that never appeared on my P&L because I never saw it leave.

Your P&L shows what came in, not what could have come in. Missed jobs aren't a line item. They just don't exist in your numbers, which is exactly why the leak keeps going.

This is also the gap CallRail can't fully close for you. CallRail tells you a call came in, what channel it came from, whether it was answered. That's valuable. It cannot tell you what your competitor billed that customer three hours later. The attribution stops at your missed call. The revenue loss doesn't.

So run the number. Write it somewhere you'll see it. If it's not alarming, you have a coverage problem you've already made peace with.

---

How I Actually Rebuilt Intake

In spring 2022, I made one rule: I answered every call personally for ninety days. The routine "can you give me a ballpark" calls at 11am. The "my breaker keeps tripping" calls at 7pm. The irritated 9:30pm callback from a customer who'd already waited two days and whose patience was gone before I said hello.

I kept a spiral notebook. Three things per missed-and-returned call: what time it came in, what the customer said when I reached them, whether they'd already called someone else. What I found was fast and not subtle.

Afternoons — 1pm to 4pm on panel-job days — were a dead zone. Friday after 4pm was its own category of lost. The voicemails that did come in were non-urgent: "just seeing if you guys are still open" is not a message that turns into a booked job by Monday. That notebook told me exactly when my phone was a liability and where I needed to build coverage.

By month nine, my NPS was 81. The answer rate drove that. Not work quality, which hadn't changed.

---

The Rationalization That Will Cost You Money

"If it's really important, they'll leave a message."

It sounds reasonable. It is wrong in residential service.

Homeowners who've never hired you, who found you on Google LSA, who have zero prior relationship with your shop — they have no loyalty to you at the moment of that first call. You're not their electrician yet. You're the first result. The second result is one tap away.

This is specifically brutal with LSA. LSA leads already close at a lower rate than organic because the customer is shopping. They saw three listings, they called you first. If you don't pick up, they move. And you already paid for that lead — Google billed you the moment they dialed — so you absorbed the acquisition cost and handed the job to your competitor for free.

You paid. Your competitor got the customer, probably got the review, and you got nothing.

I cap my LSA budget tightly for exactly this reason. If I'm not staffed to answer the calls it generates, I'm not running it at that spend level.

---

What the Intake System Actually Looks Like at Five Trucks

CallRail on every channel. Tracking numbers assigned to Google LSA, organic, the truck wraps, the Yelp page — every source gets its own number. Recording on by default, every call, no exceptions.

Thursday morning before dispatch, I sit down with my dispatcher and we listen to four calls together. Two that booked, two that didn't. We talk through what happened. The practice of actually listening is the thing — not the software, not a scoring rubric, just two people sitting down and paying attention to what's on the tape. I started doing this in 2023 and our booking rate on answered calls moved up meaningfully. I can't give you a clean before/after because I wasn't measuring it precisely enough before, but the change was not subtle.

The staffing decision underneath the software matters more than the software. My part-time CSR covers 11am to 6pm Monday through Friday, with specific attention to Friday afternoons that used to be a dead zone. Her job isn't dispatch. It's the phone.

I hired her sixty days after the bookkeeper. Not simultaneously. I tried simultaneous in early 2023 — I couldn't onboard two new people while running service, and both ended up half-trained. The bookkeeper came first, got the financial picture cleaned up, and by the time I brought on the CSR I had accurate data to actually train her with. Sequence matters more than I thought it would.

---

What to Do Monday Before You Open Your First Job

Pull the number first. Before the schedule, before anything — go into your phone system and pull missed calls for the last 30 days. Take that count, multiply by your average ticket, multiply by your booking rate, multiply by 12 to annualize. Write it down somewhere visible.

Set up one CallRail tracking number this week. Base tier. Don't let them sell you the Premium plan with AI call scoring — I ran it for a quarter and it wasn't worth it. One number, assigned to your main Google presence, recording on. Commit to sitting down with whoever handles your phones next Friday and listening to four calls together. Two that booked, two that didn't. You're not building a system yet. You're finding out what's actually happening on your line.

Block 20 minutes on your calendar right now for that Friday session. It won't happen otherwise. I skipped it three weeks running before I made it a standing appointment.

The missed calls are already in your data. You just haven't looked yet.

---

FAQ

If I'm solo or two trucks, do I really need call tracking software?

More than a five-truck shop does. At two trucks, one missed call per day represents a bigger share of your revenue opportunity than it does at scale. CallRail base tier runs $50 to $75 a month. The recording and source attribution alone justify that at any size. Start with one tracking number, recording on, and listen back once a week.

What's a good call answer rate target, and how do I find out where I am?

I run at 85%-plus during business hours and I think anything below 70% is a staffing or coverage gap, not a customer service problem. To find out where you are: pull 30 days of call data, count answered versus missed, divide. Don't use a weekly average — break it down by hour and day of week. The gap almost always clusters in specific windows, and that tells you exactly where to focus coverage.

I've tried an answering service and the quality was bad. Is there a version that actually works?

The failures I've heard about from other shops almost always come from giving the service no real guidance and expecting them to figure it out. Write an actual intake script — five or six questions, a clear path for emergency versus standard scheduling, and a hard instruction to book a callback time rather than just take a message. Ruby and Posh are the two I've heard positive things about. I haven't used either directly. Both cost real money. Neither will match a trained in-house CSR. But a scripted answering service is better than no answer.

My average ticket varies a lot by job type. How do I calculate missed call cost accurately?

Segment it. Pull your booked jobs from the last 90 days, break them out by job type, calculate an average ticket per type. Then estimate what share of your inbound calls fall into each category based on what callers typically describe. Weight the missed call cost accordingly. It takes an extra 30 minutes and the number is a lot more honest than a blended average that hides the panel upgrades.

My calls are getting answered but my booking rate is still low — intake problem or pricing problem?

Start with the recordings before you touch pricing. I've seen shops drop prices trying to fix a booking rate problem that was actually a phone manner problem. Listen for where callers are leaving the conversation. If they're gone before you even get to price, it's intake — the person answering isn't capturing urgency, isn't handling the first objection, something like that. If they're consistently leaving right after you give a number, that's a different conversation. The recording shows you which one it is. Don't guess.

When does it make sense to split CSR and dispatcher into two roles?

When your dispatcher is dropping calls because they're on the scheduling board. In my shop, the split became necessary around four trucks. Below that, one person handling both is workable. Above that, you start getting coverage gaps during busy dispatch windows because the two jobs pull in opposite directions at the worst possible times — high call volume almost always coincides with high scheduling complexity. Give each job to one person.