Your Intake Process Is Broken — And You Built It By Accident

My NPS in month two of Reeves Electric was a 4.

Not four out of five. Four out of ten. On a scale where anything below a 6 is actively damaging your reputation, I was running a business that customers were walking away from unhappy — and the technical work was fine. The panels were wired correctly. The circuits passed inspection. The trucks showed up on time.

The intake was the failure. And I hadn't built an intake. I had accumulated one, call by call, bad review by bad review, until I had a collection of habits that looked like a process if you didn't look too hard.

The Part of Your Business You've Never Actually Designed

Here's a question worth sitting with: if I asked you to write down your intake process in five sentences or fewer, could you do it?

Not describe it verbally while waving your hands. Write it. The sequence of events, the questions asked, who handles the call when you're under a panel, what happens when the customer mentions a competitor's price, what the follow-up looks like when someone says "I'll think about it."

Most owners I talk to can't do it. Because there isn't a process. There's a habit. There's what you did on the first call you took, refined slightly by the call that went badly, further adjusted by the time a customer complained, gradually calcified into something that feels intentional but was never actually designed.

The technical skills got trained formally — apprenticeship, journeyman exams, code classes. The intake got built on the fly. And nobody ever told you that the thing you built on the fly was, in practice, your primary customer-facing product.

The 90-Second Window Where You Win or Lose the Job

The booking decision is mostly made in the first 90 seconds. Not when the customer hears the price. Not when they read your reviews.

They're running a fast scan. Is this a real business? Will this be a hassle? The signals have nothing to do with technical competence. It's whether the phone got answered cleanly, whether the person who picked up sounded like they expected the call, whether the customer got put on hold before they'd finished their first sentence.

I know this because of a practice I've run for three years: every week, I sit down with my dispatcher and we listen to four calls. Two that booked. Two that didn't. We tag what went well and what didn't. That practice is what moved our booking rate from the low forties into the mid-sixties between early 2023 and the end of the year.

If you don't record your calls, you are making decisions about your intake based on your best-case memory of how it sounds — which is almost always more flattering than the reality.

I've said for years that 80% of small-shop revenue leakage happens on that first call — not in the job itself, not in the follow-up, but in the first two minutes before the customer even knows what you charge. I can't prove the exact percentage. What I can prove is that my NPS went from 4 to 81 without changing the technical work. Something changed. It was the intake.

The Ninety Days I Answered Every Call Myself

Six months into running Reeves Electric, I got a call at 9:30 on a Tuesday night from a customer whose breaker kept tripping. She wasn't furious yet — controlled, clipped, at the stage just before furious.

Except she hadn't explained it to me. She'd explained it to whoever answered my phone at 2pm when she called to book. That person — me, that day, apparently — had taken the job information, scheduled the appointment, sent the confirmation. What I had not done was note that the breaker had been tripping for three days, that her teenage son was using a space heater in his bedroom, or that this was already the second time she'd called about the same circuit.

I had planned for everything when I launched. The pricing model, the software stack, the branded vans. What I had not planned for was the fact that I was now the person on the other end of that 9:30pm call.

So for the next ninety days, I answered every call myself. Not because I had time. I absolutely didn't. I did it because I had no other way to find out what was actually happening.

What I found wasn't surprising once I saw it. Calls were being answered but jobs weren't being qualified — so we'd show up without enough information to diagnose anything before arriving. People who said "let me think about it" just disappeared, because nobody followed up. We were losing bookings we didn't know we were losing.

Month two: NPS of 4. Month nine, after rebuilding the intake from scratch: NPS of 81.

Intake Is the Customer Experience — Not a Prelude to It

Most contractors think the real product is the work. The installation, the repair, the panel upgrade, the code-compliant circuit that passes inspection. Everything before that — the call, the scheduling, the communication — gets treated as overhead. A necessary nuisance.

That's wrong, and it costs money.

For most residential customers, the path from call to invoice is the experience. The wire in the wall is proof that it happened. What they remember, what they describe to a neighbor, what goes in the review — that's the call where someone actually listened. The confirmation text that arrived on time. The dispatcher who got their name right.

A perfect panel upgrade doesn't undo a dispatcher who put someone on hold twice and never got the job address right on the first try. The homeowner will tell people the panel upgrade went fine. They will not recommend you.

When I hear "we do good work" from a small-shop owner who's struggling with growth, my first question is always the same: have you listened to your booking calls recently? Not once, as a test. Routinely. As a practice. The answer is almost always no. And the work being good is almost never the problem.

What a Deliberately Built Intake Process Actually Looks Like

Call routing. Before a human answers, where does the call go? A call that hits a generic voicemail with no timeframe for callback and no instructions loses bookings. An after-hours message that says "we return all calls before 8am the next business day — if this is an emergency, here's what to do" sets expectations and keeps trust intact.

The opening 15 seconds. There's a script for this, and it matters that it's consistent. Business name, person's name, one sentence that signals competence. Fifteen seconds. Your dispatcher should be able to say it without it sounding like they're reading.

The five-question pre-quote screen. This is where most shops skip straight to scheduling. Before I quote anything, I want five pieces of information: approximate age of the panel, whether there's been prior work on the circuit, what the customer is actually trying to accomplish (not just the symptom), whether they own the home, and what their timeline looks like. Those five questions let me triage the job before anyone drives there. They also tell me whether I'm walking into a standard service call or a conversation about a panel upgrade and a permit.

The site assessment offer. EV charger installs are where this bites me most, but any job that might involve a panel upgrade qualifies. I offer a $99 paid site assessment, credited against the job if they book. It weeds out people collecting quotes with no intention of booking. It also sets the expectation, clearly and early, that I'm not going to give them a real number without seeing the panel. That conversation is easier to have before anyone has a deposit down.

The follow-up sequence. "I'll think about it" is not a lost job. It's a job in a queue that needs a touch on Friday. A Zapier webhook from CallRail into Jobber flags unbooked calls and pushes them to a dispatcher task. Takes about an hour to set up. At five trucks, I can track exactly which of those follow-ups convert — and it's enough every week to justify the hour I spent building it.

The stack at five trucks. CallRail for tracking and recording, Jobber or Service Fusion at the core, Zapier or Make for the connective tissue, QuickBooks Online for accounting. That's it. The value isn't in accumulating software. It's in actually using what you have.

What to Do Monday Morning

First: pull your last ten unbooked calls. If you have CallRail, this takes about eight minutes. If not, pull phone records and track down numbers that called but never appeared on a work order. Listen to all ten before Wednesday. Don't evaluate during the first pass — just listen. Where did the caller go cold? Where did they stop asking questions? Where did the conversation stop being a conversation and turn into a transaction? You'll find something. It'll be uncomfortable.

Second: write your intake process down. Five sentences or fewer, from memory, without looking at anything. If you can't do it, there's your diagnosis — you don't have a process, you have accumulated behavior. If you can write it, ask yourself one question: if I handed this to a dispatcher I hired yesterday, with no additional coaching, would they execute it the way I'd want them to? If the answer is no, you have notes, not a process.

Third: schedule the weekly call review. Thirty minutes, same time every week, with whoever handles your phones. Two booked calls, two unbooked calls. Tag what worked, tag what didn't. Run it for six weeks and you'll know more about your intake than you've learned in the last three years of running your shop.

That's Monday. The redesign comes after you know what you're actually dealing with.

---

FAQ

My booking rate seems fine — how do I know if my intake is actually a problem?

"Seems fine" is doing a lot of work in that sentence. What's the actual number — calls to scheduled appointments? I've had conversations with enough five-truck shops to know most owners are guessing high. Pull three months of call volume from your phone system and compare it to three months of new job records. That gap is your starting point. If you don't know the real number, you can't evaluate whether it's fine.

I've tried scripts before and my dispatcher hates using them. How do you get buy-in?

Have your dispatcher listen to a handful of unbooked calls before you say anything about scripts. Once you hear yourself on a call that went sideways, the script stops feeling like a constraint. After that, build the opening questions together. Let them suggest the language. What you need is consistency in the structure, not word-for-word compliance. Dispatcher owns the tone, the process owns the sequence.

We get a lot of after-hours calls. Should I use an answering service or let them go to voicemail?

Voicemail kills bookings. A customer with a tripping breaker at 9pm who hits a generic voicemail has already opened Google on their second screen. Answering services are inconsistent — the good ones exist, but you need to audit the calls regularly or they drift. What I run now is a structured after-hours message with clear callback timing, plus an auto-text acknowledging the call. It's not perfect. It's dramatically better than voicemail. If your call volume justifies it, a live answering service with a tight script is the better answer — but audit the calls monthly.

What's the actual difference between intake for a first-time caller versus a repeat customer?

Repeat customers already trust you. What they need is recognition. If your dispatcher answers a repeat customer's call and treats them exactly like a new inquiry, you're telling them the relationship meant nothing. The fix is operational: your CRM should flag returning numbers before the dispatcher picks up, or at minimum within the first ten seconds. The repeat customer call can skip the intro, jump straight to the job detail, confirm they're still at the same address. Forty-five seconds instead of two minutes, and they feel like a known customer rather than a lead.

Is there a point where automating intake hurts you?

Yes, and it's a real trap. Automated confirmation texts and follow-up sequences work. The point where it goes wrong is when the customer is trying to communicate something specific — a concern about access, a question about a past job — and the automation can't receive that signal. A customer who texts back "I need to reschedule" to an automated confirmation and gets another automated confirmation anyway is worse off than if they'd never gotten the first one. Build a human review step into anything the customer can respond to. Check the response queue daily.

How do I know what questions to ask on a pre-quote call without it feeling like an interrogation?

Give the question a visible reason. "I want to make sure I send the right tech and give you an accurate window" is why you're asking about the panel age. "So we're not wasting your time on a trip if there's a code issue we can flag first" is why you're asking about prior work. When the customer understands why they're being asked, it stops feeling like an interrogation. The right five questions take about three minutes. Three minutes prevents a two-hour truck roll to a job that was never going to scope the way they described it.