Your Dispatch Order Is a Revenue Decision

The first month I ran Reeves Electric, I dispatched the same way every small shop dispatches. Who's up? Who's close? Who's already in that part of town? I had a branded van, a pricing model I'd spent three weeks building, Jobber configured, the works. What I didn't have was any logic behind the sequence of jobs I was sending people to, and I didn't know that was a problem until my month-two numbers showed up ugly in ways I couldn't immediately explain.

Wrong tech at the wrong job. Upsell conversations that never happened. A 4pm customer who'd been rescheduled once already getting my most fatigued guy. None of that showed up in my jobs-per-day count. It showed up in my NPS score — a 4, which is roughly "this company exists" as a rating — and in average ticket sizes that should have been higher and weren't.

Dispatch felt like a logistics problem. It isn't. It's a revenue decision made at 7am, usually on feel.

The 7am Gut Call Is Costing You More Than You Think

At Lonestar Electrical Services, where I ran service from 2014 to 2021, dispatch was a function. Not a task the foreman handled before the first call. A function. Someone owned it, there was a sequence logic written down, and the morning meeting was a confirmation, not a negotiation. Forty trucks meant the cost of bad sequencing was visible immediately — a mismatched crew on a complex commercial job didn't just miss an upsell, it blew a project milestone and damaged a GC relationship. The pain was legible.

At five trucks doing residential service, the pain is quieter. A bad slot costs you a review. A junior tech on a panel job costs you the upgrade conversation. None of it shows up in one dramatic line item — and because it's slow, I didn't connect the bleeding to the sequence decision that caused it until I started pulling the data myself.

When I left Lonestar, I took everything I'd learned about pricing, estimating, and field execution. I did not take the dispatch discipline, because I thought that was a big-company problem. I was wrong. The moment you have more than two techs with meaningfully different skill levels, the sequence you set determines which capability lands on which job. That's a margin decision, not a calendar decision.

Which Tech Hits Which Customer Determines Your Average Ticket

Here's how it plays out at Reeves Electric. Three meaningful skill tiers. One tech who can run a panel upgrade from diagnosis through load calc through permit pull, handle the customer conversation, and close the additional scope. Two journeymen who are solid on standard residential service but aren't going to initiate a $4,800 upgrade conversation unprompted. A newer guy who's good on the easy half-day stuff, still learning.

The 7am gut call slots people based on geography. The tech who's already on the north side gets the north-side call, regardless of what that call is. If that call is a Level 2 EV charger install on a 1972 ranch with a 100-amp Federal Pacific panel, and the tech who shows up can't have the load calc conversation, you've just turned a major scope job into a "we'll need to get back to you" that stalls in your inbox for two weeks.

The EV charger install is where I see this most clearly. The job looks like a charger install. It's actually a service upgrade conversation. If the wrong tech is standing in that garage, the conversation doesn't happen — not because the customer wouldn't have bought, but because nobody asked.

I've tracked ticket size by tech tier and slot at Reeves Electric. Complex jobs hitting my senior tech in a morning slot close additional scope at a rate that is not close to the same job hitting a mid-tier tech in any slot, or the right tech at 4pm after seven hours on the road. The gap in your shop may be different. Run the number before you assume it's not there.

The 4pm Death Spiral Is a Sequencing Problem

The 4pm death spiral is what I call the backloaded job that nobody set up to succeed. The tech's been running since 7am. He's on his fourth job. The customer has been waiting since a 2-to-4 window that's now 4:15. The tech is moving fast because he wants to get home. The diagnostic is abbreviated. He wraps up in 45 minutes, marks it complete, drives home. The customer posts a review that night.

I learned how common this was from my call recordings — specifically from calling customers back after late-slot jobs and listening to the recordings of those calls. End-of-day jobs generated a disproportionate share of friction. A customer who waited all afternoon is already annoyed before the tech knocks. A tech who's been moving since morning is not at his most thorough. The combination is predictable.

My NPS went from a 4 in month two to an 81 by month nine. A lot of that improvement came from intake process changes. But more than I expected came from fixing which jobs we put in which slots. We moved complex diagnostics to morning, protected a mid-day slot for follow-on estimates, and stopped accepting same-day EV or panel work after 1pm unless it was a pre-screened referral.

Pull last month's completed jobs from Jobber. Sort by start time. Bucket them: before 10am, 10am to 2pm, after 2pm. Calculate average ticket value per bucket by job type. If the gap between your morning and late-slot averages isn't there, fine — you don't have this problem. If it is, you're looking at your answer.

This Isn't a Software Problem Until You've Fixed the Thinking

Jobber's dispatch board is genuinely good. The drag-and-drop is clean, the tech visibility is real. I'm not arguing tooling.

The board holds the decision. It doesn't make it. I've watched shops put real money into field service software and use it to automate a sequencing logic that was garbage before the software and is now just faster garbage. If what you're putting in is "closest truck, next available," you're dispatching bad logic at scale.

Your gut after 4,000 service calls is data. Your gut after 80 is hope. A dispatcher making a sequence call at 7am without defined criteria is operating on hope — even after three years on the job. "Good at dispatch" means something different when you've defined the rules and they're executing them well versus when they've just learned to pattern-match on imperfect inputs. Those two things feel identical from the inside and aren't.

The sequencing logic has to be written down before it goes in the board. What's the skill tier requirement for each job type? Which job types get morning slots? Which don't get scheduled past noon? When does geography override skill tier? Get that on paper, agreed on, trained. Then put it in Jobber. Until then, the board organizes your improvisation.

Build the System at Intake, Not at 7am

Here's where the intake obsession pays off in dispatch: if your CSR captures job complexity, customer history, and required skill level at booking, you can pre-build tomorrow's sequence tonight. The morning dispatch becomes a confirmation — glance at the board, make two adjustments for a no-show or weather, go. The decisions were made when the information was fresh.

At Reeves Electric, every job gets tagged at intake: job type, estimated complexity (three-tier tag), required tech level. That tag travels with the job into Jobber. My dispatcher isn't looking at a blank board in the morning. She's checking whether the pre-built sequence still makes sense given what changed overnight.

The same attribution tracking I run on marketing — CallRail on every channel, UTM tags, Airtable on every source — has to run on dispatch outcomes too. Slot, tech tier, job type, ticket size, whether there was an upsell, whether it closed. That's the feedback loop that tells you if your sequence logic is working. Without it, you're guessing. Your sequence might be fine. You don't know.

Build the tracking before you build the rules. Otherwise you'll set a sequencing policy and have no way to know ninety days later whether it changed anything.

What to Do This Week

Pull last month's completed jobs. Sort by start time, bucket into three slots, calculate average ticket value per bucket. That gap — morning versus late-slot on equivalent job types — is what you need to know before you touch the dispatch board.

Then spend an hour in Jobber building job-type templates with skill-tier tags attached. Three tiers is enough: complex diagnostic, standard service, simple repair. Attach a tech-tier requirement to each. Your CSR tags the job at intake — not your dispatcher at 7am. The board has structure before the morning starts.

Last: pull four call recordings this week from jobs that started after 2pm. Listen for diagnostic depth, tech tone, whether the upsell conversation happened. That's your ground truth on whether the 4pm problem is real at your shop.

FAQ

My dispatcher has been doing this for three years and she's good. Why does this apply to us?

Good and intentional aren't the same thing. She's internalized a lot of pattern-matching — some of it right, some of it your bad habits. The question isn't whether she's good. It's whether you've written down the sequencing rules you actually want, tested them against ticket outcomes, and built them into the board. If not, she's executing her best guess at your preferences. Worth fixing regardless.

We're only three trucks. Is this worth thinking about at this size?

Three trucks is when to build the habit. You have enough volume to see patterns and not enough margin to ignore them. Setting up skill-tier tags and slot logic in Jobber takes a few hours. Waiting until five or six trucks means retrofitting the system while you're already behind.

How do I know if my 4pm jobs are actually underperforming?

Sort last month's completed jobs by start time, calculate average ticket value by slot, control for job type. If late-slot averages are significantly below morning averages on comparable work, there's your answer. Run the number.

We use Jobber. What specifically should I set up?

Start with job tags — three or four that map to complexity and required tech tier. Attach them at intake, not at dispatch. Use the scheduling requirements field to flag senior-tech-only jobs. Build two or three job-type templates with those tags pre-populated so your CSR isn't making the tagging decision fresh every time. That's the minimum. Add time-slot rules once the basics are working.

My techs push back when I change their routes. How do I handle that?

Tell them why. If a tech understands you're routing complex diagnostics to him in the morning because he's your best at closing upgrade scope — and that's reflected in his comp — the conversation changes. Pushback is usually about fairness or flexibility. Tie skill-tier assignments to pay. Give techs their schedule the night before instead of at 7am. The resistance tends to go away when people understand the logic.

When does a dedicated dispatcher make sense versus owner or CSR handling it?

Around four to five trucks running simultaneously. Below that, a sharp CSR can handle intake and scheduling with the right tools. Above that, the sequencing decisions get complex enough that splitting them from customer-facing intake is worth it. At six or seven trucks, if the owner is still touching dispatch because "nobody else knows the guys," that's a training problem.

Your Dispatch Order Is a Revenue Decision — Not a Puzzle