Freight Carriers Cannot See Which of Their Own Work Repeats, and Lose Drivers and Rates Together

Abstract

Transportation operators run two retention problems that are usually assigned to different departments. Driver turnover belongs to human resources. Shipper retention belongs to sales. The sector's own figures indicate they are one problem seen from two positions, and that the variable governing both is whether the work repeats. Less-than-truckload carriers report annualised driver turnover near 10%; truckload carriers report 66% to 106% from the same labour pool, the same equipment class and broadly the same wage band. The commercial side splits along the same line, with contract freight at 80% to 85% of truckload procurement and spot freight taking the remainder. This paper's argument is not that repeatability matters, which the figures establish on their own. It is that a carrier cannot act on repeatability because it holds no revenue operations record. Lanes, shippers and refused quotations exist as events in an inbox, a load board and an owner's memory, and none of those can be asked which relationships recur, which lapsed, and which were never pursued. The paper sets out the four records that make the question answerable, and the constraint that governs any remedy: 91% of carriers operate ten trucks or fewer, which places the enterprise answer out of reach by arithmetic rather than by preference. It draws on the logistics literature on turnover and freight pricing (Miller and Bolumole 2020; Miller and Muir 2020; Correll 2024), on carrier–shipper coordination under a spot market (Wang and Wen 2021), and on adoption constraints in freight technology (Orji and Kusi-Sarpong 2020).

The variable the sector's own figures identify

The standard account of driver turnover is a supply account: too few qualified drivers, unattractive work, carriers competing for a scarce resource. Every element is partly true, and together they predict that turnover should be broadly similar across carriers facing one labour market.

That prediction fails against the sector's own reporting (American Trucking Associations 2021). Less-than-truckload carriers report annualised turnover at approximately 10%. Truckload carriers report 66% at small fleets and 74% at large fleets, with the large-fleet series having peaked at 106% in the second quarter of 2012. A rate above 100% means a carrier replaced more drivers than it employed across the year.

Both segments recruit from the same licensed population, operate the same class of equipment and work under the same hours-of-service regime. A supply explanation cannot produce a ten-fold difference between them, because supply is common to both.

What differs is structural. Less-than-truckload consolidates at terminals and moves freight through a hub-and-spoke network, which produces repeating routes and a driver who sleeps at home. Truckload over-the-road assigns a full trailer to a destination, which produces irregular routing, rerouting that can happen daily, annual mileage above 100,000 and rest taken in the cab.

Miller and Bolumole (2020) examine industry-level truckload turnover longitudinally and find it responsive to operating conditions rather than to a static labour supply. Correll (2024) reaches the same place from driver-level operational data: the variables carrying predictive weight are features of how work is scheduled and executed.

The commercial side divides along a line of the same shape. Contract freight accounts for 80% to 85% of truckload procurement: agreed lanes, agreed rates, a shipper relationship with a history. Spot freight is the irregular form, where a load appears, a rate is quoted against current conditions, and the transaction concludes without establishing anything that persists.

Wang and Wen (2021) model carrier–shipper coordination in the presence of a spot market and show that available spot capacity changes the risk each side will carry inside a contract. Miller and Muir (2020) trace turnover through into truckload freight pricing, which relocates the problem: turnover stops being a cost absorbed internally and becomes an input to the rate a shipper sees.

So repeatability governs both what a carrier pays for drivers and what it can charge shippers. That much the published figures settle.

The question a carrier cannot answer

Knowing that repeatability governs the outcome is useless to an operator who cannot see which of their own work repeats. Ask a carrier which shippers moved freight with them on a given lane over the last year, what was quoted to each, and which quotations were refused and why. Very few can answer without reconstructing it, and reconstruction from memory is not an answer a commercial decision can be taken on.

The reason is not negligence. It is that nothing in a typical carrier's operating estate is shaped to hold the answer. A load board shows what is available now and retains nothing about the relationship. An accounting system records what was invoiced, which is the subset of work that was won, and says nothing about work that was quoted and lost. Dispatch notes describe execution rather than commercial history. Email holds everything and can be asked nothing.

Each of those systems is correct at its own job. None of them is a record of the book: one place where the commercial history of a relationship is held, and can be asked a question.

The consequence is specific. A carrier cannot pursue contract freight deliberately without knowing which shippers already produce repeat volume, because the shippers worth approaching for a contract are precisely the ones already moving freight irregularly at spot rates. That list exists in the invoices, unaggregated, alongside every one-off load, with no field distinguishing the two.

Herold and Fahimnia (2023) examine digital freight forwarders against incumbents and identify where the shipper relationship is captured. The intermediary holding the record of which shipper moved what, on which lane, at what rate, occupies the position from which a relationship can be made to repeat. A carrier without that record has surrendered the position regardless of how much of the physical work it performs.

What has to exist as a record

Four objects, and the list is deliberately short because the constraint in the next section will not tolerate a longer one.

The shipper, as a durable record rather than an invoice line. Who they are, which lanes they move, what has been quoted to them and when they last moved freight. An invoice proves a transaction happened. It is not a relationship record and cannot be asked about the ones that stopped.

The lane, as an object with a history. Origin, destination, the shippers who have moved it, the rates accepted and refused. Lanes are how a carrier's business is actually organised and almost never how its records are organised.

The quotation, with an outcome and a reason. A rate quoted and refused is the most informative record a carrier produces and the one most reliably discarded. Losing on price, on equipment availability and on transit commitment call for three different responses. Without a reason field every loss is attributed to price, which is the explanation that requires no action.

A review date on anything contracted. A contracted lane with a renewal date existing only in the owner's memory is a contract that will be renegotiated under time pressure, from whatever position the spot market happens to offer that week.

None of that requires a transportation management system. All of it is unavailable to a carrier whose commercial history lives in a spreadsheet one person maintains and nobody else can read.

Why the remedy has to be small

Any answer has to survive a structural fact. More than 55% of carriers operate a single truck, and 91% operate ten trucks or fewer.

That distribution disqualifies most of the standard technology answer. A full transportation management system installation runs at approximately $40,000, and heavy customisation repays itself only in fleets above roughly five hundred trucks. Industry guidance puts the threshold at which such a system begins to make financial sense at five or more trucks with two or more operations staff. Below that, carriers work from spreadsheets and free cloud tools. Web-based options have brought the entry price for a ten-truck fleet to approximately $1,000 a year, though the cheaper ones are frequently described as carrying dated interfaces and performance problems.

Orji and Kusi-Sarpong (2020) study adoption factors in freight logistics and find perceived complexity and available technical capability weighing heavily against adoption independently of cost. For an owner who is simultaneously the dispatcher, the salesperson and often a driver, implementation effort is not a secondary consideration. It is the binding one.

The conclusion is not that small carriers should buy enterprise software. It is that the commercial record is the part worth building first, because it is the part that converts spot exposure into contract freight (Wang and Wen 2021), and because four objects can be built anywhere while a transportation management system cannot.

Why this is not only a trucking finding

Trucking supplies the evidence because trucking publishes the comparison. Two segments of one industry, sharing a labour pool and a wage band, report turnover an order of magnitude apart, annually, through their trade association. Few sectors hand an analyst a natural experiment that clean.

The mechanism is not specific to trucks. What separates the segments is whether work is scheduled against a repeating pattern or dispatched against whatever arrived that morning, and that distinction exists in any operation assigning people and equipment to jobs. A marine dealership running service and repair has it, where work either arrives against a maintenance schedule the yard controls or as whatever is trailered in that week. A fleet maintenance operation has it. An aggregate hauler moving material to job sites has it in the most literal form, since the job sites move.

The test is identical across all of them, and it is a test about records rather than about operations. Can the operator name the customers whose work recurs, state when each is next due, and say which quoted jobs were lost and why. An operator who can answer is running a book. An operator who cannot is running a queue, and a queue is refilled from outside every time it empties.

What this argument does not establish

The turnover comparison is between segments rather than a controlled experiment. Less-than-truckload and truckload differ in more than route regularity: terminal networks require capital that favours larger and older firms, union density differs, and freight characteristics differ. The claim defended is that operating structure carries substantial explanatory weight, not that route regularity is the sole cause.

The turnover figures are annualised rates published by a trade association rather than a statistical agency, and they move with the freight cycle. The 2012 peaks reflect conditions specific to that recovery.

No causal claim is made that a commercial record produces contract freight. The relationship argued is weaker and more defensible, and it is a claim about capability: a carrier that cannot identify its own shipper history is unable to pursue contract freight systematically. Carriers win contract freight on service quality, equipment availability and price. A record is a precondition for pursuing it deliberately rather than a substitute for the rest.

The four objects described have not been measured against a control group of carriers without them. Nothing here reports an outcome at a named operator, and the figures are drawn from published industry data and peer-reviewed mechanism so that a reader can check them.

Conclusion

The freight industry's two retention problems are one problem, and the sector's own numbers say so: work that repeats holds drivers at 10% turnover and holds shippers under contract, and work that does not repeat costs a carrier its people and its rates at the same time.

That finding is only useful to an operator who can see which of their own work falls on which side of the line. Most cannot, because a load board, an accounting system and an inbox are not a record of a book, and between them they hold every fact required while making the question unanswerable.

For the 91% of carriers running ten trucks or fewer, the systems marketed as the answer are priced for fleets fifty times their size. What is actually needed is narrower and far cheaper: a shipper, a lane, a quotation with a reason on it, and a date. Four records, and the carrier can finally see which of its relationships repeat, which have lapsed, and which were never asked to.

References

Correll, D. (2024). Predicting and understanding long-haul truck driver turnover using driver-level operational data. Expert Systems with Applications, 238, 122782. https://doi.org/10.1016/j.eswa.2023.122782

Herold, D. M., & Fahimnia, B. (2023). The digital freight forwarder and the incumbent: A framework to examine disruptive potentials of digital platforms. Transportation Research Part E: Logistics and Transportation Review, 176, 103214. https://doi.org/10.1016/j.tre.2023.103214

Miller, J. W., & Bolumole, Y. (2020). Exploring longitudinal industry-level large truckload driver turnover. Journal of Business Logistics, 41(4), 294–315. https://doi.org/10.1111/jbl.12235

Miller, J. W., & Muir, W. A. (2020). The effect of truckload driver turnover on truckload freight pricing. Journal of Business Logistics, 41(4), 316–336. https://doi.org/10.1111/jbl.12252

Orji, I. J., & Kusi-Sarpong, S. (2020). Evaluating the factors that influence blockchain adoption in the freight logistics industry. Transportation Research Part E: Logistics and Transportation Review, 141, 102025. https://doi.org/10.1016/j.tre.2020.102025

Wang, X., & Wen, X. (2021). Carrier-shipper risk management and coordination in the presence of spot freight market. Transportation Research Part E: Logistics and Transportation Review, 152, 102287. https://doi.org/10.1016/j.tre.2021.102287

American Trucking Associations. (2021). Truckload turnover rate reporting. https://www.trucking.org/news-insights/truckload-turnover-rate-rises-slightly-first-quarter

American Trucking Associations. The truth about trucking turnover. https://www.trucking.org/news-insights/truth-about-trucking-turnover