For lumping agencies
Dispatching lumpers across several facilities: the scheduling problem and how agencies solve it
Once an agency staffs more than two docks, scheduling stops being a calendar and becomes an assignment problem with qualifications, travel, and fairness in it. Here is the shape of it and the practices that hold.
- Published
- Length
- 6 min read
The short version
Multi-site dispatch is an assignment problem constrained by qualification, geography, contractual commitment, crew continuity, fairness, and daily uncertainty, and it gets solved in the roster before it gets solved in the calendar. Hold qualifications with expiry dates and site familiarizations against each worker so assignment can filter rather than guess. Build the week in order: dedicated shifts with named backups, recurring known work, a deliberate float pool, then overflow against what remains. Run geography as zones, pay and rotate cross-zone travel, and keep travel projects out of the daily pool. Then fix the communication layer, because confirmed assignments and one named contact per building prevent more failures than any optimization.
On this page6 sections
An agency with one account does not have a dispatch problem. It has a schedule, and the schedule lives in the owner's head, and it works. An agency with six accounts across a metro has a genuinely hard assignment problem, and the moment it stops being solvable in someone's head is rarely the moment anyone notices. What people notice is the symptom: a crew short on a Tuesday, a worker who quit because he was sent forty minutes out three days running, a facility told a crew was coming that never arrived.
This is the anatomy of that problem, and what agencies that have grown through it actually do.
What makes it hard
The naive version is filling shifts. The real version has at least six constraints pulling against each other.
Qualification. Not every worker can work every door. Forklift evaluation by truck type, freezer experience, site-specific familiarization, badge or background clearance at facilities that require it, food safety training. A worker is qualified for a subset of your accounts, not all of them, and the subset changes as certifications lapse.
Geography. Workers commute, usually without a car they would choose to put highway miles on. Travel time is unpaid in most arrangements and it is the single largest driver of quiet attrition. A dispatcher optimizing only for coverage will produce a schedule that is perfectly filled and that half the crew will not accept twice.
Commitment shape. Dedicated On-Site Crews are promised. On-Demand / Overflow calls are not. The dedicated schedule has first claim on qualified people, and overflow gets what is left, which means a big overflow call can only be honored if the dedicated base is already covered.
Continuity. The value you sell a facility is a crew that knows the building. Every substitution spends a little of that. A dispatcher who treats workers as interchangeable is destroying the thing the account is paying for.
Fairness and hours. Who gets the good shifts, who gets the short calls, who is near overtime, who has been asked to travel most recently. Crews track this precisely even when the agency does not, and perceived unfairness costs more people than pay does.
Build the roster before you build the calendar
The recurring mistake is treating dispatch as a calendar problem and the worker roster as an address book. It is the other way around. A dispatcher can only assign well if the roster knows things.
At minimum, each worker record should hold: qualifications with expiry dates, the sites they are familiarized at, the services they are actually good at, home area, whether they drive or depend on transit, availability by day and shift, and a recent history of where they worked and how it went. Once that exists, assignment becomes filtering rather than remembering, and a second dispatcher can cover the first one's week off without the operation degrading.
The expiry dates matter more than they look. An operator whose evaluation lapsed last month is not assignable to a forklift door, and a dispatch system that does not know that will happily send him. Wiring qualification into assignment is the difference between a records problem and an incident.
Anchor, then fill
Agencies that dispatch well tend to build the week in a fixed order rather than reacting to whichever phone rang last.
Anchor the dedicated schedules first. Every Dedicated On-Site Crew shift for the week gets named people, with a named backup per shift, before anything else is committed. These are contractual promises and they are not the place for flexibility.
Then place recurring known work. Standing appointments, regular container days, the Thursday reset that happens every week. This is predictable volume that is not contractually dedicated, and it can be crewed from the same familiar pool.
Then reserve float. A deliberate pool of workers not assigned to anything, sized from your own history of no-shows and late calls. Float is what turns a no-show from a crisis into a phone call, and agencies that skip it are borrowing coverage from tomorrow every time something slips. What that looks like in practice is covered in no-show coverage.
Then take overflow calls against what is left. Which means being able to say no, or to say yes with a stated crew size smaller than asked. An overflow yes that strips a dedicated account is not revenue, it is a trade of a good account for a one-off.
The order is the discipline. It is also what lets a dispatcher answer a facility honestly and immediately instead of promising and hoping.
Geography is a routing problem, not a preference
Treat the metro as zones rather than addresses. Assign zones to workers, assign facilities to zones, and let the default assignment stay inside the zone. Cross-zone assignments then become a deliberate act with a reason, rather than the accidental output of a filtered list sorted by availability.
Two practices make cross-zone work sustainable. Pay for it, whether as travel time, a mileage allowance, or a differential, and say so plainly, because a worker who feels the travel was acknowledged accepts it again. And rotate it, so the same three people who happen to own reliable cars are not carrying every long assignment.
Project & Travel Crews are a separate category and should be run as one: a defined crew, a defined project, lodging and travel handled as part of the engagement rather than improvised, and rates that reflect it. Running travel work out of the daily dispatch pool is how agencies discover they have been subsidizing a distant account for a quarter.
The part that is communication, not scheduling
Most dispatch failures a facility actually experiences are information failures. The crew was assigned and nobody told the crew. The appointment moved and nobody told dispatch. A worker texted the crew lead at four in the morning and the crew lead was asleep.
So the operational fixes are mundane and they work.
- One channel for assignments, with confirmation. A worker is not scheduled until the worker has acknowledged. Silence is not a yes, and treating it as one is where most Monday shortfalls come from.
- One named contact per facility, both directions. Your dispatcher knows who to call at the building, and the building knows who to call at your office, including at three in the morning when the shift actually starts.
- A day-before check, not a day-of scramble. A confirmation pass the afternoon before catches most problems while there is still time to solve them.
- Same-day change rules agreed in advance. What notice a facility owes for a cancellation, what you owe for a short crew, and what happens to pay when a shift is cut. Settling this in the service agreement removes the argument from the moment when nobody has time for it.
As account count grows, this stops being feasible on a whiteboard and a phone. That is usually the point at which an agency moves dispatch into a system that holds qualifications, sites, assignments, and confirmations in one place, whether that is LumpLogix or something else. The important thing is that the roster and the schedule stop being separate objects, because the whole problem is the join between them.
- Standards
- How articles are checked
- Corrections
- team@lumplist.com