Your Intake Process Is Leaking Revenue Before the Job Starts

My NPS score in month two of running Reeves Electric was a 4. Not 40. Not 4.0. A four, out of ten, on a scale where anything below 6 means your customers are actively telling people to avoid you.

The work was good. I know that because I was doing most of it myself. No failed inspections. So I did what most shop owners do when the numbers don't match their self-image — I assumed the survey was wrong.

It wasn't wrong. The problem wasn't my work. It was everything that happened before my truck pulled into the driveway.

The Gap Nobody Measures: First Contact to Confirmed Appointment

Here's the thing about a small shop that's "doing fine." Calls get answered. Jobs get booked. The schedule fills up. That feels like a functioning intake process.

It isn't. It's a result. Results without measurements are vibes.

The actual math: calls received, calls converted to booked appointments, revenue per converted call. Most shops collapse all of that into one — "yeah, we book most of what comes in" — and that's where the measurement gap opens up.

If you don't know your booking rate by channel, you don't know your intake rate. If you don't know what a converted call is worth on average versus a lost call, you can't quantify the leak. And if you can't quantify it, you can't fix it, because you're not entirely sure it exists.

It exists.

What's Actually Happening in Those 90 Seconds

A fumbled intake call doesn't feel like a loss. The customer says "okay, I'll think about it and call back." You say "sounds good, we're here." You both hang up. Nobody yelled. Nobody left a review. The call ended without a booking, and you moved on.

But there are specific, repeating failure modes. Once you've heard them on recordings, you can't unhear them.

Hold-time drop-off: customer calls, gets put on hold for 90 seconds, hangs up. You never knew they called because the missed-call notification got buried. Scope-skip: whoever answers takes the job description at face value — "I need an outlet added in the garage" — books it at face value, nobody qualifies the panel situation or the permit requirement. And then there's what I started calling the cold-number problem — someone says a dollar amount too early without anchoring the value, the customer goes quiet, the call ends without an appointment.

None of that feels catastrophic when it happens. That's exactly the problem.

I figured this out the hard way, sitting in my truck one night with my phone and a legal pad, listening back to recorded calls from the week. I'd started using CallRail about three months in — call tracking software, records every call, attributes it to whatever channel drove it. The sales page leads with the AI scoring feature. I'd skip that for now. What I actually use is the recording library.

Once a week I pull four calls with my dispatcher. Two that booked, two that didn't. We sit there with a notepad and tag what went wrong. Forty-five minutes. No agenda except to find the moment in each call where the outcome was decided. That moment is earlier than you think.

Run It Like a Funnel, Because It Is One

Here's what I resisted for the first few months: the idea that a residential service call has stages that can be measured individually. It felt clinical. I got over it.

Call answered. Scope qualified. Appointment set. Appointment confirmed. Job starts. Leakage happens at every hand-off, not just the first call.

A high booking rate isn't proof your intake works. It might be proof that nobody on your team is asking the questions that would cause a customer to pause.

The EV charger market is where I watch this play out most clearly right now. Customer calls wanting a Level 2 charger install. If the panel has capacity and the run is short, it's a half-day job — parts and labor in the $1,100 to $1,300 range on the jobs I'm pricing in East Austin. But a lot of those calls are sitting on 100-amp service, old panels, and a garage on the opposite end of the house from the meter. The real scope is a panel upgrade, possibly a service upgrade, and the number goes from $1,200 to $4,800 to $9,200 depending on the load calc and what the utility wants.

What I see happening at shops I talk to: they quote $599 or $799 to win the call — matching what the national franchise advertises — then discover the panel scope after the deposit's down and write a change order. The customer feels bait-and-switched, even if the change order is completely legitimate. The 1-star review goes up that night. The shop blames the customer for being unreasonable.

The fix is a 20-minute pre-quote phone screen with five qualification questions, plus a $99 site assessment that gets credited back against the install if they book. That screen is intake doing its actual job — qualifying scope upfront so the appointment you're dispatching a truck to is real, not a placeholder.

Taking down a name and an address is not intake. Arriving at a confirmed, scoped, priced appointment that both parties actually understand — that's intake.

If Your Booking Rate Is Above 90%, That Should Worry You

A lot of shop owners will read this and think, "my booking rate is fine — we book almost everything that comes in."

That should make you nervous.

I've seen booking rates that high go two ways. Pricing is so low there's no competition for the job. Or — more commonly — nobody's qualifying scope, so there's nothing in the conversation that would cause a customer to hesitate. You're booking jobs without the information that would actually matter later.

A mis-scoped job doesn't look like lost revenue on your P&L. It shows up as a truck on a job that runs three hours over. A change order conversation that blows up the customer relationship. A technician who shows up frustrated because the job wasn't what he was briefed on. A review that surfaces two weeks later when you've already moved on. None of that has a line item. It just costs money in ways that don't add up on paper.

I ran into the same invisible-cost problem when I was doing my own admin work. When I hired a part-time bookkeeper and a dispatcher, I could see the payroll number. I couldn't see the 25 hours a week I'd been spending doing $25-an-hour work instead of quoting jobs — that cost was invisible, so it hadn't felt like a cost. Same thing here.

Record the calls. Find out what's actually happening. Your read on your own intake process is built on whatever sample of bookings you've thought about consciously. That's usually not a big enough sample to trust.

The 90-Day Fix

After month two and that NPS score, I did something that felt like a step backward at the time. I started answering every call myself. All of them, for ninety days. If the phone rang and I was in the truck, I pulled over and picked up. I took structured notes after every call — categories, not impressions.

What I tracked: wrong scope (customer described one job, actual job was different when the truck showed up), calls that went cold the moment a dollar amount came up without any anchoring, missed calls where I called back and found out they'd already booked someone else, and appointments where we said "see you Tuesday" but nobody confirmed and nobody got a confirmation.

That last one was the one that actually wrecked me. I had no-show rates I hadn't been tracking, because I hadn't thought of them as a measurement. When I added a confirmation step — a text that goes out 24 hours before the appointment and requires an actual reply — my no-show rate went from about one in six appointments to something like one in fourteen. One change. The customer's number was already in the system. I just hadn't been using it.

By month nine, the NPS was 81. And I want to be specific about what drove that, because it wasn't better workmanship — the workmanship hadn't changed. It wasn't faster response times — we were actually busier and in some cases slower. What changed was that customers arrived at their appointment knowing what was going to happen, what it was going to cost, and when we'd show up. That's it.

I got that wrong for the first two months of running this business. I thought customer satisfaction was about the work. It's mostly about whether what happens matches what the customer was told to expect. And that's entirely an intake problem.

Next Monday Morning: Four Things Before the Week Is Out

No overhaul. Just measurements.

Pull your actual booking rate by channel for the last 30 days. Not total calls — booked calls divided by answered calls, broken out by source. If you're on CallRail, this takes about 15 minutes. Write the number down. That's your baseline, and right now you don't have one.

Listen to four calls this week. Two that booked, two that didn't. Notepad, find the moment in each call where the outcome was decided. It's usually one exchange — one question asked or not asked, one number mentioned without being anchored, one pause that went one way instead of another. You're looking for categories of failure, not post-mortems on individual calls.

Build a confirmation step if you don't have one. A text, 24 hours before the appointment, that requires an actual response. The exact text I use: "Hey, this is Reeves Electric confirming your appointment tomorrow at 10am. Reply YES to confirm or call us to reschedule." If you don't get a response, your dispatcher calls. Build this before Friday. It's one of the highest-return single changes I made in year one.

Write down your five scope-qualification questions for your most common job type. Not a script — just the questions that, if answered, tell you whether the job is what the customer thinks it is. For EV charger installs, mine are: year of the house, current service size, panel location relative to garage, any recent electrical work, and what charger they're planning to use. Five questions. Index card. The CSR asks them on the first call. The technician shows up to a real job.

Start there. Training a dispatcher on the full qualification flow, running weekly call reviews, building a confirmation workflow in Jobber — that comes later. Get the measurement first. Everything else is a guess until you have numbers on paper.

---

FAQ

I answer every call myself and I'm pretty good on the phone — does this still apply to me?

Probably more than it applies to someone with a dispatcher. The problem with answering your own calls is you have no external check on your process. You can't hear yourself the way a customer hears you. You also tend to skip qualification steps because you've already mentally scoped the job before the customer finishes their sentence — I did this constantly in the first few months and didn't realize it until I listened back. Record your calls anyway. Listen to two or three a month. You'll find something.

What's a realistic booking rate target for a residential electrical service shop?

Honestly, I'm less interested in giving you a target than in getting you to establish your current number and track the direction it moves. I don't know your market, your pricing tier, or your job mix. What I'll say from running Reeves Electric: when our booking rate spiked above 90% for a few weeks in early 2023, it turned out we'd loosened up on scope qualification and were booking jobs we didn't actually understand yet. The number looked good. The job performance that month didn't. High booking rates and low no-show rates together are a good sign. High booking rate alone tells you less than you think.

How do I get my dispatcher or CSR to actually use a call checklist without it sounding robotic?

Don't call it a script. Give them five questions that need to be answered by the end of every call and let them figure out how to get there. Then listen to recordings with them weekly and coach on specific moments, not general performance. I had a week in 2023 where my dispatcher was letting customers skip the panel question on EV charger calls — she thought it was awkward to ask. Listening back to actual calls and talking through that specific exchange did more than any coaching conversation about "qualifying scope" would have.

I've tried recording calls before but I never actually listen to them — how do you build the habit?

Attach it to something that already happens. I do call review during my Monday dispatcher meeting — same meeting I was already having. Four calls, 45 minutes. It's on the calendar as a recurring block, not an intention. That's the whole trick. If you don't have a standing meeting to attach it to, Thursday afternoon before you close out the week works. Same time, every week, in the calendar. The review happens because it has a time slot, not because you remember to do it.

At what point does intake become a hiring problem instead of a process problem?

When the process is documented and the person executing it can't or won't follow it. CallRail gives you the data. Jobber or Service Fusion gives you the confirmation automation. But if your CSR is skipping the qualification questions because they're in a hurry, or going off-checklist because they don't like confrontation, that's not a software problem. The software will show you it's happening. That's actually the useful part — you can see it in the call recordings instead of inferring it from downstream job problems. What you do about it is a management and hiring question.

Is there a version of this that works if I'm still running solo or with one truck?

Yes, and it's simpler. At one truck, your intake funnel is you and your phone. The booking rate measurement still matters. The confirmation text still matters — maybe more, because a no-show at one truck is a bigger percentage of your day than it is at five. The five qualification questions still matter. The call recording still matters. What changes is cadence. You don't have a dispatcher to review calls with, so do it yourself — 20 minutes on a Saturday morning, two calls, a notepad. You're not building a system for a team yet. You're building the habit that the system will eventually run on.