How Trucking Carriers Run a Haul, From Tender to Settlement

A haul is not a dot moving across a map. It is a commercial agreement that becomes a plan, then a trip, then a set of documents, then an invoice and a driver settlement. Every stage has an owner. Every stage can lose a fact the next stage needs.
This is the lifecycle as carrier operations actually run it, written for people who dispatch, bill, or manage a fleet. It pairs with the problems carriers keep fighting. FleetOS sits beside this lifecycle as a decision layer on top of the TMS and telematics you already use. It does not replace the steps below.
1. Load tender
A tender is an offer: origin, destination, dates, equipment, commodity, weight, and a rate or a request for one. It arrives by EDI, a portal, or email. The first operational question is whether the offer matches equipment you run and a lane you can cover without breaking a prior promise, including home time.
Accepting in the portal and forgetting to create the order in the TMS, or the reverse, is how phantom freight starts. One system of record for the accept/decline, with the tender ID stored on the order, is the minimum.
2. Rate confirmation
The rate confirmation is the written commercial terms: linehaul, fuel program if any, accessorial rules, and free time. Dispatch and billing both need it. If the only copy is a PDF in an email thread, detention and layover will be negotiated from memory later.
Store the confirmation on the order before the truck is assigned. Note the free-time rule in fields a person can filter on, not only inside the PDF. The PDF is the evidence. The fields are how you run the day.
3. Planning
Planning decides which freight the fleet should pursue and how power and trailers should be positioned. It is where empty miles are either accepted on purpose or inherited from a hole in the board. A plan that ignores driver hours, shop status, and trailer type will look efficient until dispatch tries to execute it.
Write down the planning horizon you actually use (the rest of today, the next two days) and which constraints are allowed to break the plan. A plan nobody is allowed to override is fiction. A plan with no constraints is a wish.
4. Driver assignment
Assignment matches a load to a driver and a tractor, and often a trailer. The inputs that matter are location, remaining hours, endorsements, home-time commitments, and whether the unit is mechanically available. Hours of service are a constraint, not a suggestion the ELD will sort out.
When you reassign, record why. Patterns (repeated reassignment off one customer, or off one unit) are invisible if the board only shows the current name.
5. Dispatch
Dispatch is the live control of the plan: sending the driver, handling the check calls or the app updates, and deciding what to do when the plan breaks. The failure mode is a decision that stays in a call. “We’ll pick up tomorrow instead” has to land on the order, or the customer, the driver, and billing will each hold a different date.
6. Pickup
Pickup needs an arrival time, a departure time, and the shipper’s reference. Early arrival is not the same as loaded. If the facility holds the driver, that interval is the raw material of a detention conversation, and it has to be captured before the driver is on the road thinking about the next stop.
BOL numbers and seal numbers belong on the order at this moment. Reconstructing them from a photo three days later is how billing queues grow.
7. In-transit tracking
Tracking is location plus status. Telematics can supply location. Status (delayed, reconsigned, stopped for a breakdown) is an operational edit. Customers and brokers ask for updates. The useful system is one that answers from the order, not from whoever last spoke to the driver.
Breadcrumbs without an exception rule create noise. Define what “off plan” means for your fleet: late against the appointment window, stationary beyond a threshold you choose, or a hours-of-service conflict with the remaining miles. Review the exceptions. Do not staff a person to narrate every ping.
8. Detention
Compare arrival and departure with the free-time rule on the rate confirmation. If the stop exceeded it, create the accessorial while the timestamps are fresh, and attach the evidence the customer’s portal will ask for. Writing it off is a commercial decision. Losing it because nobody coded it is an operations failure.
9. Delivery
Delivery repeats pickup’s discipline: times, references, exceptions, and OS&D notes if something is short or damaged. A clean delivery status with a damaged product is not a clean delivery. The note has to exist before the customer files a claim from their side.
10. Proof of delivery
The POD closes the physical move. It is a signed document, an electronic acceptance, or the consignee’s portal confirmation, whichever the customer defined. Until it is attached to the order, billing should not treat the load as complete. Chasing PODs is cheaper on the day of delivery than at day 15 of the aging report.
11. Billing
Billing assembles the invoice from the rate confirmation, the accessorials, and the POD. Mismatched references are the usual rejection reason. A “ready to bill” status that requires those pieces, and a queue that lists the missing one, keeps the delay visible. Month-end is a poor time to discover a week of loads that never had a POD.
12. Settlement
Driver settlement pays the move according to your pay rules: per mile, percentage, stops, extras. It depends on the same miles and accessorials billing used, or the driver and the customer will be living in two versions of the trip. Disputes about pay are often disputes about a detention or a layover that was never coded, not disputes about the pay formula.
13. Cost allocation
Revenue is on the load. Some costs are on the load (contracted accessorials you paid out, a lumper). Some are on the truck for the week (fuel, a repair). Some are on the fleet (insurance, plates). A contribution view has to say which bucket a dollar falls into. If fuel is allocated by gallons from the card and miles from telematics, write that rule down and apply it the same way every week.
This is the layer FleetOS is built to make explicit: one cost-per-mile and margin model operations and finance can both open. The TMS remains the order system. The model is the agreement about money.
14. Profitability analysis
After allocation, you can ask which customers, lanes, and assets covered the costs you defined. Keep the question tied to that definition. A lane that looks profitable before deadhead and shop time can look different once those are attached. Changing the definition mid-quarter makes the comparison useless. Change it on purpose, document it, and start a new period.
15. KPI monitoring
A short set is enough if each KPI has an owner and a source:
- On-time pickup and delivery, using the appointment you stored, not a reconstructed one.
- Empty miles, separated into planned repositioning and unplanned deadhead.
- Detention events created versus stops that exceeded free time.
- Loads delivered but not ready to bill, and why.
- Units dispatched with an open downing work order.
- Contribution after the written allocation rules, by the slice you manage (customer, lane, or fleet).
If a KPI does not change a decision this week, it is a report, not a control. Park it.
16. How data moves between systems
The tender might start in a portal. The order lives in the TMS. Location and hours live in the ELD. The shop lives in a maintenance system. Fuel lives with the card provider. Billing might push to an accounting package. Each hop can drop a field. The practical integration is not “connect everything.” It is to name the field that must survive each hop: tender ID, appointment, arrival, departure, document status, and the pay and bill codes.
17. Where operational data gets lost
- Acceptances that never become orders, or orders with no tender ID.
- Rate rules trapped in a PDF.
- Reassignments and delay decisions that stay on the phone.
- Arrival and departure times that are never written down.
- Accessorials remembered after the driver has delivered three more loads.
- PODs and receipts in a personal camera roll.
- Fuel and repairs that are never tied back to a unit and a week.
Lost data is why the problem list feels permanent. The lifecycle above is also a checklist of capture points. You do not have to instrument all of them in one project.
18. Automation opportunities
Automate the comparisons a person currently does by eye, and leave the commercial exceptions to a person.
- Create an exception when a stop’s dwell exceeds the free-time field.
- Hold billing until POD and required references are present.
- Notify dispatch when telematics or the shop marks a dispatched unit unavailable.
- Match fuel transactions to a tractor and flag anything unmatched.
- Produce the morning list of loads off plan, instead of a static board everyone re-sorts by hand.
Those are the same shapes described in the automation ROI guide: a trigger you already store, a queue, and residual work for the cases the rule should not decide.
19. Fleet-level analytics
Fleet-level analytics is the lifecycle rolled up without changing the definitions. Morning review is a list of exceptions and a contribution snapshot leadership can reconcile to the TMS and the fuel bill. It is not a second set of numbers from a spreadsheet that “adjusts” the TMS until it looks right.
That is the operating idea behind FleetOS: ingest the systems you have, validate before the warehouse is polluted, and publish a morning brief and exception queues. Analytics that cannot name their source fields will restart the argument the haul already created.
20. Operational checklist
- Tender accepted or declined once, with the ID on the order.
- Rate confirmation stored, free time captured as data.
- Plan states which constraints can be broken and by whom.
- Assignment checks hours, home time, and shop status.
- Dispatch decisions written back to the order.
- Pickup and delivery times captured at the stop.
- Detention coded while the timestamps are fresh.
- POD attached before ready-to-bill.
- Settlement uses the same extras billing used.
- Allocation rules written down and left stable for the period.
- A short KPI list, each with an owner.
- A weekly look at which capture point is still leaking.
Conclusion
The haul is a closed loop from offer to cash and driver pay. Carriers already own software for pieces of it. The leaks are at the handoffs: a PDF instead of a field, a phone call instead of a status, a receipt that never reaches billing. Close one handoff, measure it, and move to the next. A fleet intelligence layer is justified when those handoffs are captured and the remaining problem is that nobody shares a definition of the result.
Frequently asked questions
Where should we start if the whole lifecycle feels leaky?
Start at ready-to-bill or at detention, because both have a document you can count. Pick the capture point that already creates aging invoices or unpaid waiting time, and make that status impossible to skip. Expand only after that queue is real.
Does FleetOS replace the TMS or the ELD?
No. Those systems remain the records for orders and for hours and location. FleetOS is the governed layer that joins them to maintenance and fuel so contribution and exceptions use one definition.
How is this different from a broker’s workflow?
A carrier owns the assets, the drivers’ hours, the shop, and the settlement. A broker’s workflow is commercial coverage without that asset loop. Mixing the two checklists hides the costs that only show up on the fleet.
Haseeb Tariq
Co-Founder at Vyntrix Labs
Keep reading

Operational Problems Trucking Carriers Still Have to Manage
Empty miles, detention, maintenance, dispatch, billing, and fragmented TMS and ELD data, why these carrier problems persist, and where a shared operating picture changes the work.
Sufyan
AI Engineer

FleetOS: Why Fleet Intelligence Is More Than Dashboards
Motor carriers already own TMS, telematics, and shop systems. FleetOS is the governed decision layer on top, cost-per-mile truth, exception-first ops, and AI that explains what to do next.
Ashir Haroon
CEO
How to Measure the ROI of Workflow Automation
A practical way to estimate labor time, errors, delay, and payback for a workflow automation project, using formulas and labeled hypothetical numbers, not borrowed case-study results.
Haseeb Tariq
Co-Founder