When should tomorrow's dispatch actually be built?

The afternoon before, not the morning of. A schedule built at 6 a.m. is being built while crews are already standing in the yard waiting for it, which converts every problem into an emergency. Building it the previous afternoon leaves a window to solve equipment shortages and crew gaps while people are still reachable.

Who should own dispatch?

One person, with authority to say no. Dispatch fails most often not from bad information but from divided authority — a salesperson promising a crew, an owner reassigning one, and a dispatcher discovering both at 7 a.m. The role needs the power to decline work that does not fit, subject to escalation.

How do you handle emergency callouts without wrecking the schedule?

Decide in advance what capacity is held for them rather than deciding in the moment. Companies that do meaningful emergency work generally reserve a crew or accept that a specific low-priority job gets bumped. The failure mode is treating each callout as a one-off crisis and cannibalizing whichever job is least able to complain.

What should the morning release actually contain?

Job, site, arrival time, crew, the device list, the permit or plan reference, the customer contact, and anything unusual about access. If a crew has to call the office to find out where they are going or what they need, the release was incomplete.

How do you know if dispatch is working?

Track how many jobs change after the release goes out. A small number is normal and healthy. A large number means the schedule is being built on information that is not yet firm, and the fix is upstream in intake and confirmation rather than in dispatch itself.

The Daily Dispatch Routine for a Traffic Control Company

What a working dispatch day actually looks like — the afternoon-before build, the morning release, the midday reality check, and the end-of-day close — plus the three habits that separate companies that dispatch well from those that firefight.

Dispatch is the part of a traffic control company where every other function's failures become visible. Bad intake shows up as a crew standing at a site that is not ready. Bad equipment records show up as a truck missing an arrow board. Bad estimating shows up as two flaggers scheduled for a job that needs three. Dispatch does not cause any of that, but it absorbs all of it.

Which is why "we need better dispatch" often turns out to mean something else. Still, the routine itself matters, and there is a shape that works.

The afternoon-before build

The single most important structural choice is when tomorrow gets built. Companies that build it the previous afternoon operate differently from companies that build it at 6 a.m., and the difference is not effort — it is the availability of solutions.

At 3 p.m. the day before, a missing arrow board is a phone call to a rental yard. At 6:15 a.m. the next morning, it is a job that starts late. The problems are identical; only the options differ.

A workable afternoon-before sequence:

Confirm tomorrow's jobs are real. Customer confirmed, site accessible, permit in hand, no weather cancellation pending. A job on the board that nobody has confirmed since Monday is a rumor.

Assign crews and check the arithmetic. Right number of people, right certifications, hours that do not create an overtime or rest problem you did not intend.

That last check has a regulatory edge worth respecting. Where drivers operate commercial motor vehicles, hours-of-service rules published by the Federal Motor Carrier Safety Administration constrain what you can legitimately schedule, and applicability depends on vehicle and operation specifics — see FMCSA requirements for traffic control fleets, and confirm your own status with qualified advice rather than by analogy to another company. Separately, flagger assignment has to respect training expectations; OSHA's flagging provisions sit at 29 CFR 1926.201 alongside the broader employer material on its highway work zones page. A dispatch board that does not know who is certified will eventually send someone who is not.

Reserve equipment against actual availability. Not against what you own — against what is in the yard and not already committed. This requires knowing where deployed devices are, which is exactly the record most companies do not have. See tracking traffic control devices by job site.

Resolve the conflicts you just found. This is the whole point of doing it early. There will be two or three most days.

The morning release

The release is the artifact crews actually work from, and its quality determines how many phone calls the office fields before nine.

A complete release carries: the job and customer, the site with enough detail to actually find it, arrival time, crew assignment, the device list, the permit or plan reference, a customer site contact, and anything unusual — gate codes, escort requirements, restricted hours, adjacent work.

Two things worth being strict about.

It goes out at a predictable time. Crews plan around consistency more than around early. A release that lands at 5:30 some days and 7:15 others trains people to ignore it and call instead.

It is one document, not a conversation. If half the information is in the release and half was mentioned verbally to a foreman, you have no release. The verbal half evaporates and the written half gets blamed.

For crews, the release is also the input to the tailgate briefing and should carry enough about the setup and hazards to support one.

The midday reality check

Somewhere between 10 a.m. and noon, someone should establish what is actually true: which crews are on site, which jobs have started, what has changed.

This is the step most commonly skipped, and skipping it is how a company discovers at 4 p.m. that a crew has been waiting since eight for a site that was not ready. The check does not need to be elaborate — a status per active job is enough. What matters is that it happens on a schedule rather than when someone happens to wonder.

The midday check is also when tomorrow's problems first become visible. A job running long today is tomorrow's crew shortage, and finding that at 11 a.m. gives you the afternoon to solve it.

The end-of-day close

Closing the day is where the money is, and it is the step most often deferred to "when things calm down."

Three things have to happen:

Tickets come in. Signed, complete, matched to jobs. A ticket that arrives Thursday for Monday's work is a ticket the customer has had three days to forget the details of — see preventing daily ticket disputes.

Equipment movement is recorded. What went out, what came back, what stayed on site. This is the record that makes rental billing possible and makes lost devices traceable to a job rather than to a mystery.

Exceptions are noted while they are fresh. Extra hours, work outside scope, a device damaged, a customer request. Each of these is either captured today or argued about in six weeks — see change orders on traffic control jobs.

The practical constraint is that all three depend on a crew doing something at the end of a long day. Which means the capture has to be fast enough that doing it is easier than not doing it. A form that takes four minutes on a phone at the site gets filled in; a form that requires a stop at the office does not.

Emergency callouts

Companies that do incident response, utility emergencies, or after-hours work face a structural tension: the schedule assumes stability and the callout destroys it.

The answer is to decide the policy before the phone rings. Either hold capacity — a crew or a person on call, with the cost accepted as the price of the service — or establish an explicit bump order so that when a callout lands, everyone already knows which job yields. What does not work is deciding in the moment, because the job that gets bumped is then whichever customer is least likely to complain, which over time is a customer-selection strategy nobody chose. See emergency callout dispatch.

The three habits that actually separate companies

One schedule, visible to everyone who needs it. Not a schedule plus a whiteboard plus what the owner knows. Most dispatch failures are two-versions-of-the-truth failures.

A written record of changes, not just of plans. When a job moves, the fact that it moved and why is worth more later than the original plan. It is also how you find out that a particular customer changes dates constantly, which is a pricing conversation.

Dispatch authority that is real. If the dispatcher cannot decline work, they are not dispatching, they are transcribing — and the schedule will be overcommitted by whoever talked to a customer most recently.

Where the routine breaks under growth

The routine above works by hand up to roughly three to five crews. Past that, the binding constraint is not effort but concurrency: several people need the current answer simultaneously, and the paper or spreadsheet version develops versions. That threshold, and what changes on either side of it, is covered in scaling a traffic control company.

Traffic OS is built around this daily shape — dispatch, crews, and equipment on one board, releases that carry the device list and plan reference, and GPS-stamped signed tickets that close the day from the field. Pricing is flat by company size rather than per user, which matters here specifically: dispatch only works if every crew is actually in the system, and per-seat pricing gives you a reason to leave the seasonal ones out. If you want to see it against your own week, book a walkthrough.