Driving is overhead, and nobody pays you to sit at a light. Most bad routes happen because jobs get booked in the order the phone rang, so geography never gets a vote. Order the day geographically instead of chronologically: group the stops that sit near each other and run the day as a loop rather than crossing town twice. It only works if every job carries the real property address and an honest duration, and it is worth more again if the client gets a heads-up when you are actually on the way.
Look at last Tuesday honestly.
Five service calls. The first one northeast, because that client phoned on Monday morning. The second one across the river to the southwest, because that client phoned Monday afternoon and "tomorrow around ten" came out of your mouth before you thought about it. The third back northeast, four blocks from the first. The fourth downtown. The fifth back southwest again, a twelve-minute drive from stop two.
You crossed the city twice, drove past one job site on the way to another, and got home forty minutes later than the work itself justified. Nobody planned that route. It planned itself, one phone call at a time, in the order the phone happened to ring.
That is how most small-crew days get built, and it is worth being precise about what it costs, because the fix is nearly free.
Why does drive time cost you money?
Every hour of your working day is one of three things: billed work, work that wins billed work (quoting, follow-ups), or overhead. Driving is overhead. Necessary overhead: the work is at the client's property, not yours. But overhead all the same. Nobody pays you to sit at a light.
Put a number on it. If your working rate is $95 an hour, an hour of driving is $95 of capacity you spent moving a van. Add fuel at, say, $18 for the extra distance a badly ordered day adds, plus the wear that mileage quietly banks against you. A crew of two doubles the labor side, because both of you are in the van.
The uncomfortable part is that drive time hides. A late materials delivery is visible and annoying. Forty minutes of avoidable driving is spread across a day in seven-minute slices, and it feels like work: you are moving, you are between jobs, you are "on the road." It never gets counted, so it never gets fixed.
The other cost is softer but real. The day's last client gets the tired version of you, and the buffer that would have absorbed a surprise (a seized valve, a talkative client, a supplier stop) was spent on pavement instead.
The drive cost calculator shows what the drive to a job costs you, per job and across a year.
Why do days get booked in the wrong order?
No contractor sits down and chooses a bad route. Bad routes are the default output of a reasonable process:
Jobs are booked in the order clients call. The calendar fills chronologically by phone call, and geography never gets a vote. The Monday-morning caller gets the 8 a.m. slot regardless of where they live.
Times get promised on the phone. "Around ten" is easy to say and hard to walk back. Once three clients hold loose time promises, the order is frozen before anyone looks at a map.
The plan lives in your head. With five stops there are 120 possible orders. Nobody mentally compares 120 orders while making coffee at 6:40 a.m. You pick something plausible and drive.
The day changes after it starts. A cancellation, an emergency call, a part that has to be picked up. Now the order you half-planned is stale, and you are re-planning at a red light.
None of these are discipline failures. They are what happens when route order is a side effect instead of a decision.
How much time does route order actually save?
Take a plausible service day in a mid-sized metro: five stops, spread the way real calls spread.
Booked in call order, the day runs: shop → northeast (14 km) → southwest (19 km) → back northeast (17 km) → downtown (8 km) → southwest again (12 km) → shop (11 km). Total: 81 km, and with city traffic call it roughly 1 hour 50 minutes behind the wheel.
Reordered as a loop (the two northeast stops together, then downtown, then the two southwest stops on the way home), the same five jobs run about 52 km and roughly 1 hour 10 minutes of driving.
Same clients, same work, same van. About 29 km and 40 minutes handed back, on a day that was not unusually bad. Run a service-heavy week like that and the difference is on the order of two to three hours. That is enough for one more small job, or for getting home before dinner every night instead of twice a week.
The point of the example is not the exact figures; your city, your stops and your traffic will produce different ones. The point is that the savings come from order alone. No one drove faster, no job was cut short, no client was dropped. The waste was purely structural, which is exactly the kind of waste software is good at removing.
The Day Board: stops you can reorder in a tap
This is the job of the Day tab on Zeus's Work Board. The Work Board has three views: a Dispatch lane board for who is doing what, a multi-day Gantt Planner for how jobs stack across the week, and the Day view. The Day view takes today's stops and does the thing you were doing in your head at 6:40 a.m. It puts them on a map, and one button orders them. One practical note, and this is the honest place to say it: the Work Board is a Pro feature. Job scheduling itself is on the free plan — the three boards are what the paid one adds.
The mechanics matter less than what they replace, so here is the honest shape of it:
Every stop comes from work you already scheduled. There is no separate "route planning" data entry. Jobs carry their property address; the Day view reads today's scheduled work and lays it out. If the address is on the job, the stop is on the map.
The order is optimized, not chronological. One button reorders today's stops around geography instead of around who phoned first: the loop version of the day, not the ping-pong version. You see it as an ordered list of stops and as pins on a map, so a wrong-feeling order is visible before you start the engine, not at the second river crossing.
ETAs are realistic rather than optimistic. Each stop shows when you will plausibly arrive given the order and the driving between stops. That number is the difference between "around ten" and a promise you can actually keep. It updates your own expectations too, which is how you stop booking six stops into a five-stop day.
The day survives contact with reality. When a client cancels at 8:15 or an urgent call lands at noon, you are not re-solving the puzzle on a phone keyboard at a light. Drop the stop, add the new one, run the ordering again, and what is left comes back in a sensible sequence.
What this changes in practice is small and constant: you stop making routing decisions entirely. The 6:40 a.m. staring-at-the-calendar ritual becomes opening the Day view, sanity-checking the order, and driving.
"On the way," without the phone call
The second half of a well-run service day is what the client experiences, and the biggest complaint clients have about trades is not price or even quality. It is silence. "They said morning" curdles into a one-star mood by 11:15, even when you arrive at 11:20 having done nothing wrong.
The traditional fix is calling ahead, which is a good habit that dies under load: you are driving, your hands are full, the number is in a text thread somewhere.
Zeus builds the habit into the route without pretending to be a messaging service, and the distinction is worth stating plainly. From the stop you are about to drive to, Notify Client writes the message for you (their name, your ETA, where you are heading) and hands it to your own Messages app or your mail, addressed and ready. You press send. Nothing goes out through Zeus, and nothing goes out on its own. What it removes is the part that actually kills the habit: composing a sentence and hunting for a number while you are trying to leave. The client who was watching the window relaxes; the client who nipped out to the store comes home. You have converted the worst part of their day, the uncertain wait, into the feeling of a delivery they can see moving.
If you prefer to text or call personally, the script barely matters as long as it happens:
"Hi, it's Sam from the plumbing company. I'm wrapping up my current job and heading your way, should be there about 11:20. Anything changed since we spoke?"
The last question earns its place. Roughly one time in ten it surfaces something (a gate code, a dog, an "actually while you're here") that is far better discovered in the van than in the driveway.
The inputs that make the route honest
An optimized route is only as good as what it knows. Three habits keep it truthful, and all three are things you should be doing anyway:
Put the real property address on the job. Not "the Hendersons," not the billing address for a rental the client owns across town. The job's address is where the van goes; if it is right, everything downstream (the map pin, the order, the ETA) is right for free.
Give jobs honest durations. A "quick filter swap" that is actually ninety minutes with travel up and down a ladder poisons every ETA after it. You do not need precision; you need to stop scheduling in wishes.
Anchor what is genuinely fixed. Some stops are truly time-bound: a condo building that only allows work 9 to 11, a client who must let you in before their shift. Those are anchors, and the rest of the day should flex around them. The failure mode is treating every casual "around ten" as an anchor; then nothing can move and you are back to the zigzag by another name.
And because the Day view is part of an offline-first app, the route you loaded in the morning is still on your phone in the parkade with no signal. That story is told properly elsewhere on this blog; here it is enough to say the plan does not evaporate when the bars do.
What the route can't know
A field guide should be honest about limits, so: route optimization orders your stops well. It does not read minds.
It does not know that Mrs. Kowalski talks for twenty minutes every visit, that the downtown stop has no parking after 3 p.m., or that you always grab lunch near the east-side supplier. It does not know about the crash that just closed the bridge. It measures distance as the crow flies rather than by road, so a river or a rail cutting between two nearby pins can make a tidy-looking order worse than it reads. A stop whose address it cannot place on the map is not ordered at all; it goes to the end of the list until you fix the address. And it will occasionally propose an order that is geometrically right and humanly wrong: putting the messy demolition stop before the client who will notice a dusty crew walking through a finished home.
That is what your judgment is for, and why the ordered list is a starting point you can override, not a sealed verdict. The software does not plan a perfect day. The win is that you start from a good day and adjust, instead of starting from a random day and firefighting. Ten seconds of review beats ten minutes of mental routing, every morning, forever.
Frequently asked questions
Is this worth it if I only have two or three stops a day?
The routing win shrinks with fewer stops. With two stops there are only two possible orders. But the rest of the Day view still earns its place: realistic ETAs, the on-the-way notify, and the address and job details in one place instead of across a calendar, a text thread and your memory. The habit also scales with you; the crew that grows into six-stop days does not have to build a new system on the day it needs one.
What about jobs that run multiple days? Do they clutter the route?
Multi-day work is what the Gantt Planner view is for; it shows how those jobs occupy the week. The Day view is at its best on service-shaped days (several shorter stops) and on the days when a long-running job shares the calendar with a couple of side calls, which is exactly when routing mistakes get expensive.
My clients expect exact appointment times, not windows. Does an optimized order break that?
No. It makes the times you promise more likely to be true. The order is decided the evening before or at the start of the day, and each stop's ETA comes from that order. The discipline that changes is when you promise: quote a time after the day has an order, or give a window on the phone and firm it up with the on-the-way notify. Promising exact times at booking, days ahead, was always the fragile part.
Does the route order account for traffic?
Treat the ETAs as planning numbers, not live navigation. The Day view is about the order of stops and a realistic sense of when you will reach each one; your maps app of choice still handles turn-by-turn and the live traffic on the way to the next pin. The two jobs are complementary: one plans the day, the other drives the segment.
ZEUS