The Peter Principle — Laurence J. Peter’s 1969 observation that in a hierarchy every employee tends to rise to their level of incompetence — has found many applications in the world of work over half a century. In the last two years, though, a new factor has emerged that deserves serious discussion: AI and automation are changing the manager’s job, and this change has direct effects on how and where incompetence shows up.
The thesis of this article, in one line: well-designed fleet software doesn’t replace the fleet manager — it raises the level at which they become incompetent, because it frees them from the work that drowns them and gives them time for the work that truly requires managerial skills.
That’s a strong claim. Let’s see why it holds.
The average fleet manager spends time the wrong way
If you ask a fleet manager at a transport SMB what they did in the last eight hours, the most common answer includes:
- checking tachograph downloads and handling compliance alerts
- re-planning two or three operational routes after disruptions (customer unavailable, traffic, breakdown)
- updating a spreadsheet with the day’s delivery data
- answering three or four driver phone calls on operational issues
- preparing a report for the operations director on the week’s anomalies
Out of eight hours of workday, on average six or seven are spent on tracking, reporting, and tactical scheduling. Only one or two hours — on good days — are left for what truly requires managerial skills: talking to a driver in difficulty, deciding whether to add a sixth person to the shift, building a plan to reduce turnover, understanding why the most important customer is showing signs of dissatisfaction.
This imbalance is the heart of the problem. The work that fills the fleet manager’s days is exactly what an excellent individual contributor — a logistics veteran, an experienced ex-dispatcher — naturally excels at. And it’s exactly what doesn’t require managerial skills.
When a driver promoted to fleet manager spends 90% of their time doing work they did better in the previous role, the 10% of truly managerial work inevitably gets sacrificed. Not out of bad will — out of pure operational saturation.
It’s the mechanism the Peter Principle has been describing for decades. What’s changing today is the availability of tools to shift the balance.
What AI handles well (and why it frees managerial time)
A modern fleet management system reliably automates several classes of activity. Not all of them. But the right ones.
Dynamic planning and re-planning. Building the optimal route plan for the next morning, factoring in volumes, time windows, access restrictions, vehicle capacity, and each driver’s remaining driving time, is a combinatorial problem an algorithm solves in minutes. Re-planning during the day when something unexpected happens is even more typically algorithmic. On this topic, the move from Excel to automated planning is the most relevant transition phase for transport SMBs.
Compliance tracking and alerts. Monitoring tachograph data, flagging driving anomalies, reminding about document deadlines, generating regulatory reports required by EU legislation: these are activities where human error is high and the value of manual control is low. A system does it better, faster, and traceably.
Standard reporting. Building the weekly fleet KPI report, populating the operations director’s dashboard, feeding the management system with delivery data: all activities that have informational value for the reader, but zero added value in doing them manually.
Routine operational communication. Customer notifications about delivery windows, completion confirmations, digital signature capture at the delivery point: digital Proof of Delivery eliminates calls and paper handling that today weighs on drivers and dispatchers.
Across all these activities, the difference between “manual fleet manager work” and “supervising an automatic system” is enormous in terms of time freed. Conservatively: two or three hours per day of manual work eliminated, out of an eight-hour day.
What AI doesn’t handle (and why the manager matters more than ever)
The opposite mistake is thinking AI solves everything. It doesn’t. There are classes of problems where the machine barely helps or doesn’t help at all, and where irreducible human skills are needed.
Work organization decisions. How many drivers do I need? Which shift should I open? Should I hire or rely on an owner-operator? These are decisions that require context, risk assessment, reading weak market signals. An algorithm can propose scenarios, but the final choice stays human.
People management. Understanding why a driver’s mood is changing, reading dissatisfaction signals before resignation, handling a conflict between two colleagues, giving constructive feedback, building trust: all of this is and stays pure managerial work. No software does coaching. No dashboard replaces a hard conversation done well.
Relationship with the strategic customer. The customer worth 30% of revenue doesn’t talk to an app. They talk to a person, and that person needs decision-making authority, relational sensitivity, the ability to represent the company. It’s work the fleet manager can do only if they’re not simultaneously trying to download tachograph data and re-plan the Padua route.
Team development. Building the supervisor you don’t have today. Identifying the next coordinator. Designing career paths that avoid the problem of failed promotions. These are long-horizon activities that generate strategic value but require continuous time, not half-hour chunks between emergencies.
AI, in other words, frees the fleet manager from non-managerial work to give them back time to spend on the work that actually is. It’s a change in mix, not in function.
The consequence: the Peter Principle shows up higher
Which brings us to the opening thesis. If the fleet manager no longer spends six hours a day on operational-tactical work but can dedicate four to real managerial decisions, two interrelated things happen.
On one hand, candidates who became “incompetent” because they drowned in tracking become potentially effective, because tracking is now handled by a system. People who failed in the manual-heavy role can succeed in the manual-light one.
On the other hand, the managerial skill level required goes up. When 50% of the time is genuinely people management, strategic data reading, and key customer relationships, the skills required are more sophisticated, not less. The “Marco the driver” who used to survive in the do-everything fleet manager role because they spent half the time still doing operations finds themselves exposed in a role where operations are automated.
The Peter Principle, in other words, doesn’t disappear. It shifts higher. The incompetence threshold rises, but the skills needed to stay below it become rarer and more valuable.
What this means if you’re planning to adopt fleet software
For a transport SMB owner evaluating the adoption of a fleet management platform, these considerations have three practical implications.
First, recalibrate your expectations about people. Don’t expect the software to turn a mediocre fleet manager into a good one. It will free them from part of the work that was sinking them, but it will more clearly expose their managerial gaps if there are any. If you have doubts about your current fleet manager, introducing the software will make them more visible, not less.
Second, rethink the profile of the next fleet manager you hire. In a company where systems handle planning, tracking, and reporting, the fleet manager progressively becomes a people manager with technical-operational skills, not an experienced dispatcher with some coordination responsibility. The skills to look for change: empathy, communication, team-building, aggregate data reading for decisions — no longer operational speed and encyclopedic route knowledge.
Third, consider AI as an investment in role attractiveness. The new generations entering the sector don’t want to spend eight hours in front of a spreadsheet. A fleet manager with modern tools spends time on work perceived as “real”: talking with people, reading data for decisions, building relationships. It’s a change that makes the role more attractive to young, qualified candidates in a labor market where the shortage of professional drivers is structural and growing.
Three useful questions to know where you stand
Three test questions for anyone running a fleet today and wondering if this discussion really applies to them.
-
Does your fleet manager spend more time on tracking-reporting-scheduling or on decisions requiring judgment? If the answer is “the former,” software can free up valuable time. If it’s already “the latter,” you’re facing the more sophisticated problem of selection and development of managerial skills.
-
In the last 12 months, how many times has your fleet manager had a development conversation with a driver (about growing, what’s working, what to improve)? If the answer is “rarely” or “never,” it’s likely not a problem of intent but of time: there’s no room in the days.
-
If you hired a new fleet manager tomorrow, would you look more in the operational logistics world or in the people management world? The right answer depends on your level of automation and the type of role you want to design. Without a conscious choice, you’ll end up hiring “someone like the previous one” — which is the surest way to repeat the previous problems.
The Peter Principle doesn’t disappear with technology. But with the right technology, and with awareness of what’s changing in the fleet manager’s role, you can move the incompetence threshold up one floor. That’s not nothing — in many cases, it’s the difference between a company that grows and one that stays put.