Your Field Software Knows More Than You Do — Use It

I had Jobber running on day one. Branded vans in the driveway, a pricing model I'd spent three months building. I had the software stack. What I didn't have was any idea what to do with the data it was quietly collecting while I was busy answering 9:30pm calls from customers whose breakers kept tripping.

That was February 2022. My NPS by month two was a 4. I was doing everything the "launch your shop" content said to do, and the signals that would have told me exactly where the wheels were coming off were sitting in my dashboard, unread, like gauges on a car I never looked at.

Here's what I should have been watching: booking rate by call source, job cycle time, and first-call resolution rate. If you're running Jobber or Service Fusion on five trucks and you can't pull those numbers right now, you've already paid for the answer — you just haven't looked.

---

Why Your Software Vendor Won't Tell You This

Field software companies are not in the business of improving your weekly review habits. They're in the business of selling seats, tiers, and integrations. The onboarding call is designed to get you activated — not to get you to a meaningful operational outcome six months later.

Owners who describe their platform as "where we do scheduling" are probably using a fraction of what they're already paying for. The vendor's onboarding doesn't correct this. They want you logging jobs. Whether you're reading the data those jobs generate is your problem.

The gap between what your field software captures and what you actually act on is not a technology problem. It's a habit problem. Your vendor sells seats. Closing that gap is on you.

I ran into this with CallRail. They have a Premium tier with AI call scoring — flags tone, intent, whether opportunities got missed. I tried it for a quarter. It wasn't worth the upgrade, at least not yet. The point of CallRail isn't the AI interpretation. It's the recording. I'd rather listen to four calls a week myself and actually understand what's happening than pay for a summary while I stay detached from the pattern.

The upsell is always more features. The discipline is always fewer metrics, reviewed more often.

---

Booking Rate by Call Source: The Number Most Shops Have Never Pulled

Your field software logs every job request. Your call tracking logs every inbound call. Connect those two things — even loosely — and you can calculate booking rate by lead source. That's the number that tells you whether your marketing spend is working or just spending.

Most owners I talk to say something like "most of our work comes from Google." That's a guess. The version that actually lets you make decisions sounds different: what percentage of calls from each source actually booked, and what did those calls cost you? I can pull that because I run tracking numbers on every channel through CallRail, tag every source with UTM parameters on digital and dedicated numbers on print and truck wraps, and track outcomes in an Airtable I built specifically to cross-reference source against booked revenue — I set it up after month two when I realized I had no idea which of my launch channels was actually working.

The reason source-level booking rate matters more than your overall rate: aggregated numbers hide problems. I've talked to shops — guys I've met through The Backcharge and at trade events — who were running what looked like a reasonable overall booking rate. When they broke it down by source, their organic Google traffic was booking well above average and their LSA spend was dragging the whole number down. They'd never seen it because they'd never segmented it.

This is the Faustian bargain with Google LSA I keep coming back to. The volume is real. But those leads close lower because the customer has seen three other listings before they called you. If you're not segmenting by source, you'll never see LSA pulling your average down. You'll see a number that seems acceptable and keep spending.

How to pull it: most field software platforms let you filter jobs by lead source if you've been tagging them at intake. If yours won't do it natively, export the last 90 days to a spreadsheet. Two columns — lead source, booked or not. Calculate the rate for each source. You'll probably see one channel that's significantly below the others. That's the one to look at first.

---

The Ninety-Day Listening Period That Rebuilt Reeves Electric's Intake

Month two, I started answering every call myself. Not because I had a system — because my NPS was a 4 and something was obviously broken.

What I found: the problems weren't the work. The work was fine. The problems were in the first 90 seconds of every call. We were quoting scope before we understood the job. We were giving price ranges so wide they sounded like we were guessing. Calls were ending without a confirmed next step. The customer hung up confused, called someone else, and occasionally left a review that made my stomach drop.

Here's what made me feel like an idiot: the software had been logging call duration, missed calls, and booking outcomes the entire time. Not because I set it up that way — because that's what it does by default. The data wasn't missing. I just hadn't been reading it.

By month three I had a practice. Once a week, I sat down with whoever was handling dispatch and we pulled four calls — two that booked, two that didn't. We listened. We noted what worked: confident scope framing, a clear next step before hanging up. We noted what didn't: long silences, pricing language that trailed off, calls that ended with "let me think about it" and nothing else. Same doc, every week.

By month nine, NPS was 81. That practice, not any single hire or any marketing change, doubled our booking rate in 2023. The software was doing its job the whole time. I just hadn't started doing mine.

---

Job Cycle Time and First-Call Resolution: The Two Clocks Your Software Is Already Running

I spent seven years at Lonestar Electrical Services running service jobs on a 40-truck operation. You learn something at that scale: cycle time is a cost center. Every day a job sits open between first contact and invoice close is overhead.

On a five-truck residential shop the numbers are smaller but the principle is identical. Your field software timestamps everything — job created, dispatched, tech checked in, invoice sent, paid. Job cycle time, call to invoice close, is sitting in your data right now.

Here's what it tells you. A shop where average cycle time is creeping up is not, in most cases, a capacity problem. In my experience it's usually a dispatch sequencing issue — jobs going to the wrong tech — or a parts problem where nobody's stocking what the job actually needs. The software is logging every timestamp that would prove which one. A tech who's consistently running long cycles isn't necessarily slower. He might just be getting dispatched the jobs that require parts nobody has on the truck. You won't know until you pull the timestamps by technician and look.

First-call resolution rate is related but different. It's the percentage of jobs resolved on the first visit without a return trip. A lot of residential shops don't track this explicitly — they absorb the cost of the second trip as a normal part of operations. It's not normal. A return trip costs you real labor time, fuel, and rescheduled dispatch. Your software tracks whether a job was closed and then reopened. That's your first-call resolution rate. It's already there.

---

What to Actually Do Monday Morning

Pull the booking rate report, segmented by lead source, for the last 90 days. Go to wherever jobs are stored in your field software and filter by lead source. If it won't segment natively, export to a spreadsheet and do it manually — two columns, source and outcome. Look for the channel that's significantly below your overall average. That's the one you're probably overinvesting in without knowing it.

Set a weekly block — same day, same time, every week. Pull booking rate by source, average job cycle time, and first-call resolution rate. Write them in the same document every single week. Don't analyze the snapshot. Watch the direction. A shop that sees its numbers moving the wrong way six weeks before it shows up in the P&L is in a different position than one that finds out at the monthly close.

Schedule your first four-call listening session this week. Pull two calls that booked and two that didn't from the last 30 days. Listen to all four. Take notes on the first 90 seconds. Don't build a scoring system yet. Just listen. The pattern will be obvious, and you'll start hearing it mid-call once you've trained your ear on it — which is when it actually changes how your team performs.

None of this requires new software. It requires the habit.

---

FAQ

I already have Jobber — does it actually track all of these metrics, or do I need a different tool?

Jobber tracks enough to get started on all three. Job cycle time is derivable from the timestamps already logged on every job. Booking rate by source requires that you've been tagging lead sources at intake — if you haven't, start now and you'll have usable data in 60 days. First-call resolution takes a little manual work unless you've been using job tags consistently. You don't need a different tool. You need to use the one you have more deliberately. CallRail layered on top of Jobber gets you the call-level data to make booking rate by source reliable — that's roughly $50-$100 a month, not a platform swap.

How do I calculate first-call resolution rate if my software doesn't have a field for it?

Look for jobs that were marked complete and then reopened, or jobs that generated a second work order within 14 days for the same address and a similar symptom. It's not a perfect filter, but it'll tell you enough to act on. In Jobber, run a client history report and look for repeat visits within short windows. It's manual the first time. Once you know the pattern, you can build a tagging habit into your dispatch process that makes it automatic.

My booking rate looks fine overall — why does breaking it down by source matter?

Because "fine overall" is usually a blended average hiding a bad channel. If referrals are booking well above average and your paid LSA spend is booking significantly below it, your overall number might look acceptable — but you're subsidizing the bad channel with the good one without knowing it. Segmenting tells you where to put more money and where to pull back. An overall booking rate feels like information but doesn't give you anything to act on. Booking rate by source is a decision-making tool.

What's a realistic benchmark for these numbers on a five-truck residential electrical shop?

I'll tell you what I see at Reeves Electric and what the guys I've compared notes with seem to run, but your market and job mix will move these around. We keep our overall booking rate above 55%, with organic and referral sources running meaningfully higher than paid. Cycle time on standard residential service calls we try to keep under three days call-to-invoice. First-call resolution we track against 80% as the target. These aren't numbers from a report — they're what I've seen work. Take them as a starting point, not a standard.

I've tried reviewing call recordings before and it never stuck — how do you actually build the habit?

Attach it to something already on your calendar. I do it Thursday morning, same block I use for the numbers review. Four calls maximum — two booked, two not. Forty-five minutes with a timer. The goal isn't comprehensive auditing. It's pattern recognition. Once you've done it four weeks in a row, you'll start hearing the pattern mid-call, which is when it actually changes how your team performs.

At what point does it make sense to hire someone to own this process?

When you can't reliably do the weekly review yourself and the shop is big enough that the gap is costing you real money. At five trucks, with the revenue we're running, that question is real — but I'd hold off on the hire until you've done the review yourself long enough to know what you're looking for. If you don't understand what the numbers mean, you can't manage someone whose job is to watch them. Get fluent in the data first. Then hire someone to run the process.