In this article (11 sections)
A trucking company with forty trucks almost always looks the same from the inside: a telematics system that came bundled with the hardware, three or four spreadsheets only one person knows how to drive, a group chat where the real dispatching happens, and fuel receipts in a folder.
None of that is wrong. Each piece solved a real problem when it showed up. The problem is that none of them talks to the others, and the consequence is that nobody in the company can answer simple questions: what that load actually cost, which truck is the worst performer, how much got overspent last quarter.
This piece is about what a fleet management system really needs, what order to build it in, and the mistakes that make these projects drag.
The opening mistake: asking for the super-app
The first conversation almost always starts the same way: we need a system that handles loads, fuel, maintenance, tires, drivers, compliance documents, customers, invoicing and payroll, with an app for the driver and a dashboard for management.
All of that is reasonable and all of it is needed. But asking for it in one piece has three predictable consequences:
- The budget lands on a number the company won't approve.
- The timeline goes past a year.
- And —most important— nothing goes into production until everything is ready, so the company spends a year paying and getting no benefit.
Fleet systems that fail almost never fail for technical reasons. They fail because they were planned whole and ran out of fuel along the way. The antidote is to pick one module that pays for itself, get it running, and fund the next one with what it saved.
Where to start
The load as the unit of costing
If you only get to pick one thing, this is it: every load should be a record with all of its costs attached.
It sounds basic and it's the piece that almost never exists. Without it, the company knows what it billed and what it spent for the month, but not what it made on each load. And without that you can't tell which customer is profitable, which lane is worth running, or whether the rate you're charging covers what the run costs.
A useful load record has: truck, driver, origin and destination, miles, dispatch and delivery dates, freight, customer, the rate billed, and the costs charged against it —fuel, tolls, driver pay and per diem, and a prorated share of maintenance.
With that alone, management moves from arguing about impressions to arguing about numbers. It's the module with the best return of any of them.
Fuel: where the money is
In a fleet, fuel is the biggest line item after payroll, and it's where the most money leaks without anyone noticing.
What a system has to be able to do:
- Record every fill with truck, gallons, amount, odometer and who fueled.
- Calculate fuel economy by truck and by driver, in miles per gallon.
- Compare against each truck's own baseline. A fleet-wide average is useless: a loaded tractor-trailer and a light run don't burn the same. What matters is how far each truck has drifted from itself.
- Flag it when something goes out of range.
This is where the real anomalies show up: fills that don't line up with the miles run, fuel economy collapsing on one specific truck —usually a genuine mechanical problem, not theft—, and gaps between what left the tank and what made it into the vehicles.
That last point deserves a note: if the company has its own fuel island, tank reconciliation is a separate module and it isn't trivial. It involves stick and meter readings, temperature-related shrinkage, and calibration. Treat it as its own project rather than assuming it's included in "fuel control".
Preventive maintenance
The point isn't to log the repairs you already did —any spreadsheet does that— but to get ahead of them on mileage.
The system has to know the PM interval for each service on each truck, know the current odometer, and warn you before it's due. Put that next to the cost history per truck and you get the information behind the most expensive decision a fleet makes: when a truck stopped being worth keeping and has to be replaced.
A fleet that moves from breakdown repairs to preventive maintenance lowers its cost per mile measurably. That's a benefit you can calculate before the module is built.
Compliance documents with expiration dates
Annual DOT inspections, insurance certificates, CDL and medical card renewals, registrations and permits, hazmat endorsements where they apply. It's the simplest module to build and it avoids the dumbest cost there is: a truck sitting idle or an out-of-service violation over a piece of paper nobody checked.
What can wait
- The driver app. It's the most requested item and almost never the first thing to build. In stage one the office can enter the load data from what already comes in through other channels. The app earns its place when you want the driver entering data on the spot, photographing the bill of lading and capturing the signature at delivery on the phone — and for that, the office process has to be running smoothly first.
- The telematics integration. It sounds essential and rarely is in the first stage. Pulling odometer readings automatically is convenient, but typing them in once per load works too. On top of that, every telematics vendor has its own API with its own quality, and that integration is a classic source of delays.
- Invoicing. If you already invoice through an accounting system that works, don't replace it. Having the fleet system feed data to the accounting system is a better idea than building invoicing from scratch.
- Executive dashboards. They come after there's data inside. A dashboard on an empty database is decoration.
How it connects to what you already have
This point decides whether the project succeeds more than the feature list does.
Don't replace the telematics. The hardware is installed and it works. The management system consumes its data, it doesn't substitute for it.
Don't replace the accounting system. It's a regulated product that gets updated as rules change. Having the two talk is enough.
Do replace the spreadsheets. That's the whole point. And there's a condition for it to happen: as long as the spreadsheet still exists, people will keep using it. A fleet system that lives alongside the old spreadsheet for months never gets adopted.
Expect the group chat to stay. The real dispatching is going to keep happening there for a good while, and that's fine. What the system has to achieve is that the outcome of that coordination ends up recorded — not that the conversation moves.
The three mistakes that stretch these projects
1. Modeling the exception before the rule. Every trucking company has its special cases: the customer with the negotiated rate, the load that gets billed differently, the driver who gets paid another way. If the system tries to cover them all from the start, the complexity multiplies. In the first version the exception gets handled by hand; you automate it once it's been shown to come up often.
2. Starting with what management likes. The dashboard with charts is the best thing to show off and the last thing to build, because it depends on everything else already feeding it data.
3. Not deciding who enters each piece of data. A fleet system lives or dies on the quality of the data going in. If nobody is explicitly responsible for entering the loads and the fuel fills, the system fills up halfway and the conclusions coming out of it are worse than having nothing. This isn't a software problem: it's an organizational decision that has to be made before the first line of code.
The numbers
For a fleet of twenty to fifty trucks, reference ranges:
- Stage one —loads with costing, fuel with economy tracking, compliance documents with expirations, a web admin panel—: USD 8,000 to 15,000, six to ten weeks.
- Stage two —preventive maintenance, a driver app with data entry and photos, telematics integration—: USD 10,000 to 20,000 on top.
- Infrastructure: on the order of USD 50 to 150 a month.
- Maintenance: 15% to 20% a year on what was built.
Against that, the calculation that matters: in a fleet of forty trucks, a fuel economy improvement of just 3% usually adds up, over a year, to a number in the same range as all of stage one. That's the math worth doing before you decide, and you can do it with the data already sitting in your fuel receipts.
Where to start tomorrow
Before you ask for a quote, pull together three things: last quarter's fuel consumption by truck, last month's list of loads with what was billed on each, and the last big breakdown repair invoice.
With that you'll be able to calculate what not having the system costs you today. And that number, not enthusiasm, is what should decide the size of the project.