The old rule and why it fails
The rule most brokerages inherit is two calls a day: one in the morning to confirm the driver is rolling, one in the afternoon to confirm they still are. It is easy to staff and easy to audit. It also fails in a predictable place. A truck that was fine at 9 a.m. and fine at 3 p.m. can have missed a pickup appointment at 11, sat at a shipper for four hours, and left with a delivery window it can no longer make, and the afternoon call will still get a cheerful “all good.”
The deeper problem is that a fixed cadence spends the same effort on every hour of a load, and the hours are not the same. A load has two moments where a bad answer costs real money, the arrival at each stop, and long stretches where the only useful information is that the truck is still moving in the right direction. Two calls a day over-checks the highway and under-checks the docks.
A cadence by phase, not by clock
Before pickup. The risk is that nobody shows up. A driver who has not accepted the dispatch, or accepted it and is still 300 miles away two hours before the appointment, is the single most expensive thing to discover late, because a missed pickup means rebooking at a spot rate on the day. Check acceptance the evening before and position at roughly the point where the drive time to the shipper equals the time left before the window. If the two do not fit, that is the call to make now.
At the shipper. The questions change to arrival, loading, and departure. Arrival time against the appointment decides whether detention will be owed. Departure time against arrival decides how much. Between the two, the only thing worth a check is whether loading has actually started, because a truck that has been on the yard for two hours without backing to a door is a phone call to the shipper, not to the driver.
In transit. This is where two calls a day wastes most of its effort. What matters is whether the truck is moving, whether it is on a road that leads to the receiver, and whether the drive time left fits inside the delivery window with the driver's remaining hours. A position every few minutes answers all three without a conversation, and a driver who has to answer the phone on the interstate is a driver you have made less safe to prove nothing has changed.
Approaching delivery. Attention goes back up. The receiver needs an ETA they can plan a dock crew around, and a late truck needs a rescheduled appointment before the window closes, not after. A useful rule: from the moment the ETA is within two hours of the window edge, either side, someone should be looking at that load every fifteen minutes or so. This is the stretch where a call to the receiver changes an outcome.
At the receiver. Same as the shipper: arrival, unloading, departure, and then the one thing that is easy to forget, the proof of delivery. A load that left the receiver an hour ago with no POD on file is a check, because the driver is easiest to reach before the next load starts.
The cadence in one table
These intervals are a reasoned default, argued through with a working dispatcher rather than measured across the industry. A reefer of produce with a tight receiver deserves tighter numbers. The rows a board can run on its own (the acceptance reminder, the in-transit thresholds, the stop clock) are settings per organisation in LoadSignal; the rest is guidance for the person. Revisit all of it after a month of real loads.
| Phase | What can go wrong | Check |
|---|---|---|
| Booked, not yet accepted | Driver never confirms | A reminder text at 10 minutes without acceptance; a person after an hour |
| Evening before pickup | Truck too far away to make it | Position once; drive time against the window |
| Morning of pickup | Late for the appointment | Position at the point where drive time equals time left, then every 30 minutes |
| At the shipper | Detention, loading not started | Arrival, then a look at 60 minutes if not loading, then departure |
| In transit | Stopped, off route, tracking lost | Position every 5 minutes; a person only when a threshold trips |
| Within 2 hours of the window | Late, receiver unprepared | ETA every 15 minutes; call the receiver if it slips past the window |
| At the receiver | Detention, missing POD | Arrival, departure, POD within the hour |
What to ask at each check
A check call that asks “how's it going” gets “fine.” The question should have a number in the answer. Before pickup: where are you now and what time will you be at the shipper. At a stop: what time did you arrive, are you at a door, what time did you leave. In transit, if a call is needed at all: what exit are you at. Approaching delivery: what time will you be there, and does that still work for you with your hours. The point is not the number itself. It is that a specific question gets a specific answer, and a specific answer can be checked against the board.
Text where you can. A driver reads a text at the next stop and cannot take a call on the road. A text with the load number and the one question in it also leaves a record on the load, which a phone call does not.
How often the customer hears from you
Less often than you check, and on a different trigger. A customer does not want a position; they want to know whether the plan changed. Three updates cover most freight: picked up, with the actual time; an ETA the evening before delivery or the morning of, whichever comes first; and delivered, with the time and the POD. Everything else is on request or on exception. When a delivery window slips, the customer hears it from you before it slips, with a new time and a reason. That single update is worth more than a week of “in transit, on schedule.”
A tracking link the customer can open changes this further. When the position and ETA are one click away, most where-is-it emails are never written, and the updates you send are the ones that carry a decision.
What changes once loads track themselves
With a position arriving every few minutes from the driver's phone, the carrier's ELD or a share link, most of the table above stops being a task. Acceptance is a status. Position against the window is a number the board computes. Arrival and departure are recorded by a geofence rather than remembered by a driver. Stopped, off route and lost are thresholds. What remains for a person is the handful of calls that change something: the shipper who has not started loading, the receiver whose window will not hold, the driver whose hours do not fit the plan.
LoadSignal checks active loads on a five-minute schedule and on relevant load events, raising exceptions against the thresholds your team sets. The real-time load visibility guide covers where the positions come from, and the playbook for a truck that goes quiet covers what happens when they stop. If you are still on two calls a day, how the check call itself gets automated is the place to start.