Trucking Industry: Why 25% of Miles Run Empty
Twenty-Five Percent Empty
One in four trucks traversing the American interstate system carries nothing. The trailer is sealed, the diesel burns, the driver collects wages, yet the cargo bay holds only air. This statistic, twenty-five percent of all trucking miles run empty, represents not merely inefficiency but the structural signature of an industry organized around fragmentation rather than coordination. There are over one million trucking companies in the United States. Ninety percent operate six trucks or fewer; ninety-seven percent operate fewer than twenty. The typical carrier is not a fleet manager dispatching hundreds of vehicles from a central hub, but a husband-and-wife team, or a father and son, often operating from a garage in a residential neighborhood much like the one in Ravenna, Seattle, where Dan Lewis’s grandfather started a distribution company decades ago.
This fragmentation has persisted because trucking resists consolidation. The capital requirements for entry are low, a used tractor and a trailer, and the operational flexibility of small scale allows carriers to survive market volatility that would break larger fleets. The result is a bazaar of independent operators who have historically coordinated through telephone calls, truck-stop conversations, and the accumulated trust of personal relationships. For decades, the industry has accepted empty miles as the inevitable cost of matching dispersed supply with dispersed demand without a centralized coordinating mechanism.
<pConvoy enters this landscape with a precise economic wager: that the fragmentation which makes trucking resistant to corporate consolidation creates the conditions for aggregation. The company’s engine runs on the systematic replacement of relational coordination, the manual work of freight brokers who match shippers with carriers through phone calls and black books, with algorithmic matching that learns driver preferences, location, and capacity in real time. By aggregating the long tail of these small carriers onto a unified platform, Convoy aims to collapse the search costs that produce those empty miles, capturing value by taking a percentage of transactions while simultaneously lowering friction costs for both sides. The model compounds as scale increases: more drivers improve matching efficiency, which reduces shipper search time, which increases volume, which improves the data that optimizes routes. In theory, the platform becomes more valuable to every participant as it grows, a network effect built atop the very fragmentation that has long prevented the industry from consolidating into a few large carriers.Yet this economic engine depends upon a peculiar alchemy. The platform must persuade independent operators, men and women who often remember the names of their brokers’ children and rely on handshake deals for credit, to trust a digital interface with their livelihoods. It must standardize interactions across a million disparate actors without destroying the personal trust that governs payment flexibility, dispute resolution, and the adaptive capacity necessary in an inherently volatile business. And it must do this while managing $668 million in venture capital, the highest valuation of any startup in the Pacific Northwest, in an industry notorious for cyclical devastation. The question is not whether the technology can optimize a route, but whether the sociology of American trucking will permit the technology to achieve the scale required to survive the inevitable cold season.
The Relationship Tax
Traditional freight brokerage operates through a tax on attention and relationship. A broker maintains thousands of personal connections with small carriers, fielding phone calls at odd hours to match available trucks with available loads. When a shipper needs capacity, the broker searches through mental inventories and phone trees; when a driver needs a backhaul to avoid running empty, the broker negotiates rates based on personal history and immediate market conditions. This coordination is manual, expensive, and fundamentally unscalable. The broker captures margin, often substantial, because the search costs for both sides are prohibitively high without an intermediary who has invested years in building trust across dispersed participants.
Convoy’s structural advantage lies in the automation of this coordination layer, but the implications extend beyond simple efficiency. Because the platform replaces the manual maintenance of relationships with data-driven matching, its cost-per-transaction declines as volume increases. Traditional brokers face fixed or rising costs; each new relationship requires additional human attention, phone calls, and social capital. Convoy, by contrast, amortizes the cost of its technology across millions of transactions, creating a widening cost advantage as it scales. This is not merely software improving an old process. It is a fundamental restructuring of the economics of trust in the industry. Where brokers charge for the privilege of accessing their curated networks, Convoy offers market access without the relationship overhead, allowing the company to operate at lower cost-per-transaction while still improving outcomes for both drivers and shippers.
The danger, however, is that relationships in trucking are not merely inefficiencies waiting to be optimized. They serve functional purposes that algorithms struggle to replicate: credit extension during cash-flow crunches, flexibility when a driver’s child falls ill, forgiveness when a delivery runs late due to weather, and the willingness to bend rules for reliable operators facing temporary hardship. The personal broker acts as a shock absorber in a volatile industry, absorbing risk through social capital rather than contractual terms. By replacing this with algorithmic matching, Convoy risks optimizing away the very flexibility that makes small-scale trucking viable for independent operators. The platform must achieve the efficiency of standardization while somehow preserving the adaptive capacity of personal trust: tension that defines every strategic decision the company makes.
The Standardization Paradox
The central constraint shaping Convoy’s trajectory is the paradox of standardizing the unstandardizable. The platform’s value depends on density: enough drivers in enough locations to ensure that a shipper can find capacity in minutes rather than hours, and enough load volume to ensure that drivers can find backhauls without the empty miles that destroy their margins. Achieving this density requires standardization, tandardized data fields, standardized payment terms, standardized expectations about pickup times, communication protocols, and route acceptance. Yet the supply side of this market consists precisely of actors who have resisted standardization. These are family businesses operating on irregular hours, handshake deals, and the kind of operational flexibility that allows a father-son team to take a day off for a baseball game without updating a software dashboard.
If Convoy moves too aggressively to standardize, demanding real-time GPS tracking, strict adherence to routing algorithms, or rigid payment schedules, it alienates the very fragmented supply base it depends upon. These drivers can revert to traditional brokers who offer the flexibility to refuse a load, to negotiate by voice, to explain circumstances that algorithms cannot parse. The mom-and-pop carrier who built a business through personal relationships may view platform demands for transparency as surveillance and standardization as rigidity. If Convoy moves too slowly, however, failing to achieve the density that makes the platform more efficient than a phone call, it collapses into irrelevance. The network effect requires critical mass, but critical mass requires convincing independent operators to surrender autonomy to a system.
This tension is not merely a go-to-market challenge; it is existential to the business model. Convoy must convince drivers that the platform offers better outcomes, higher utilization, faster payment, more convenient loads, without appearing to extract value through forced optimization. The algorithm must feel like an enabler of choice rather than a constraint upon it. Yet as the platform scales, the temptation to optimize relentlessly grows stronger. The mathematics of venture-scale returns push toward maximum utilization, toward algorithms that treat drivers as fungible capacity units rather than independent business owners. Managing this tension requires operational discipline that few technology companies possess, particularly those accustomed to the relative homogeneity of software markets or consumer delivery networks.
The Asset-Light Boundary
Beyond the immediate tension of standardization lies a deeper strategic fork: whether to remain a pure marketplace or to vertically integrate into asset ownership. The pure marketplace model, connecting shippers with independent carriers and taking a transaction fee, scales faster and requires less
capital and carries less operational risk, but it also generates lower margins and surrenders control over the final-mile experience. When a shipment arrives damaged, late, or with a driver who lacks professional conduct, the shipper blames Convoy regardless of whether the company owns the truck. The marketplace model relies on reputation systems and algorithmic matching to enforce quality, yet in trucking, quality failures can be catastrophic: missed delivery halts a manufacturing line, a refrigerated load spoils, an accident triggers liability claims that ripple through insurance markets. Without direct control over vehicle maintenance, driver training, or safety protocols, Convoy must govern a million dispersed actors through software alone.
Vertical integration offers an escape from this constraint. If Convoy owned trucks and employed drivers, it could guarantee capacity, control service quality, and capture the full margin between shipper rates and operating costs rather than merely a transaction fee. The model would shift from marketplace to asset-heavy carrier, resembling the integrated logistics giants that dominate parcel delivery. Yet this path demands a different organization entirely: capital intensity spikes, cyclical exposure increases, and the company must master the operational complexities of fleet maintenance, driver employment law, and asset utilization in downturns. More critically, vertical integration destroys the network effect. A company operating ten thousand trucks cannot leverage a million independent carriers; the platform becomes merely another competitor in a fragmented market rather than the aggregator of that fragmentation.
The tension between these models is not merely strategic but existential. Convoy must decide whether it is a technology company that moves information or a transportation company that moves freight. The choice determines capital requirements, talent acquisition, and organizational culture. A marketplace hires software engineers and product managers; an asset-heavy carrier hires mechanics, dispatchers, and compliance officers. The hybrid approach, attempting to own some assets while maintaining a marketplace, creates confusion for both shippers and drivers. Shippers wonder whether they are buying capacity or software; drivers wonder whether they are competing against the platform or partnering with it. Dan Lewis’s background at Amazon, where he learned to write six-page documents securing funding for new initiatives, suggests an organizational discipline that might navigate this ambiguity, yet the pressure of $668 million in venture capital creates its own gravitational pull toward growth at the expense of strategic clarity.
The Data Moat and the Illusion of Control
Convoy’s most durable advantage lies not in its brand or its capital, but in a data layer that competitors cannot easily replicate. The platform captures real-time location, capacity, and behavioral preferences from a fragmented supply base that has never been digitized. Traditional brokers operate through phone calls and memory; their knowledge of driver whereabouts and preferences exists in individual heads, not databases. Convoy, by contrast, builds a proprietary map of American trucking capacity that improves with every transaction. It knows which drivers prefer overnight hauls versus daytime runs, which routes offer acceptable profit margins after fuel costs, and where capacity shortages emerge before the broader market recognizes them. This information asymmetry compounds over time: more drivers generate better matching algorithms, which attract more shippers, which generate more loads, which attract more drivers.
The durability of this advantage stems from the difficulty of bootstrapping such a network. A new entrant faces the cold-start problem: without drivers, it cannot offer shippers reliable capacity; without shippers, it cannot offer drivers consistent loads. Convoy’s four-and-a-half-year head start creates a data moat that widens with scale. Competitors hiring pure software engineers cannot simply code their way into this market; they must convince independent truckers to adopt an unproven platform with no existing liquidity, while simultaneously persuading shippers to trust capacity that may not materialize. The switching costs for drivers increase as the platform learns their preferences, like a navigation system that remembers favorite routes, making departure psychologically and economically costly even if alternatives exist.
Yet this data advantage masks a fundamental fragility. Convoy controls information but not assets. It can optimize routes, suggest loads, and predict availability, but it cannot compel a driver to accept a shipment, maintain a vehicle, or adhere to safety standards. When a Convoy-matched truck breaks down on an interstate, or a driver falsifies hours-of-service logs to meet a delivery window, the platform bears the reputational cost while possessing limited remedial power. The data moat provides matching efficiency but not operational reliability. A shipper choosing between Convoy and a traditional asset-based carrier weighs the platform’s convenience against the carrier’s direct control over equipment and personnel. In stable markets, efficiency wins; in volatile markets, control prevails.
This illusion of control extends to the optimization algorithms themselves. The platform’s value proposition depends on reducing empty miles through better matching, yet drivers may resist algorithmic routing that conflicts with their autonomy. A driver who has operated the I-5 corridor for twenty years knows which truck stops offer safe parking, which shippers load quickly versus slowly, and which receivers tolerate minor delays. An algorithm optimizing for miles per gallon or asset utilization may route that driver through unfamiliar territory, to unfamiliar facilities, with unfamiliar risks. If drivers perceive the platform as extracting value through forced efficiency, maximizing their utilization without regard for their preferences, they will revert to broker relationships that offer choice. The data moat protects Convoy from competitors but not from the defection of its own supply base.
Capital as a Liability
The $668 million sitting on Convoy’s balance sheet represents not just a resource but a structural vulnerability. In an industry governed by cyclicality, capital abundance creates the temptation to overexpand during frothy markets, locking in cost structures that become fatal when freight demand collapses. Dan Lewis has explicitly acknowledged this danger, citing his experience through the 2001 and 2008 downturns as formative lessons in capital discipline. Yet the institutional imperative of venture funding pushes in the opposite direction. Investors who contributed to the highest valuation in the Pacific Northwest expect growth trajectories that justify their entry price. The pressure to deploy capital, to hire aggressively, expand geographically, subsidize rates to capture market share, conflicts with the cyclical realities of trucking, where a recession can reduce freight volumes by thirty percent overnight.
The specific risk lies in fixed cost accumulation. If Convoy uses its capital to build technology infrastructure, hire close to a thousand employees, and establish operational density in major markets, it creates a burn rate that requires sustained high volumes to sustain. When the cycle turns, and in trucking, cycles always turn, shippers reduce tender volumes, spot rates collapse, and drivers park trucks rather than accept unprofitable loads. A marketplace model with minimal overhead can contract gracefully; a capital-heavy organization must choose between deep layoffs that destroy organizational knowledge or continued cash burn that risks insolvency. Lewis’s awareness of this dynamic, rooted in his memory of previous recessions, functions as an economic variable in its own right, potentially constraining the reckless expansion that destroys less experienced competitors. Yet awareness and institutional pressure are not the same thing. The boardroom dynamics of a highly valued startup may override the caution of a CEO who remembers 2008.
Moreover, capital abundance attracts competition. The visibility of Convoy’s funding signals to markets that digital freight brokerage offers venture-scale returns, inviting well-capitalized entrants, mazon itself, traditional logistics giants with balance sheets dwarfing Convoy’s, or international platforms with patient capital. The money that enables rapid scaling also advertises the opportunity, potentially compressing margins before network effects achieve lock-in. In trucking, where barriers to entry remain low and switching costs remain uncertain, capital serves as both shield and target.
The Operator as Algorithm
Dan Lewis’s specific background, consulting at Oliver Wyman on supply chain logistics for the Panama Canal and Boeing, followed by Amazon, followed by the founding of Wavii, represents a hybrid operational expertise rare in technology entrepreneurship. Most software founders approach logistics as an information problem to be solved through elegant code; most logistics veterans approach technology as a tool for incremental efficiency. Lewis occupies the narrow intersection: he understands the physics of moving heavy objects through space and time, and he understands the economics of platform aggregation. This hybrid identity allows Convoy to avoid the typical failure modes of logistics technology startups, which either underestimate the operational complexity of trucking or overestimate the willingness of truckers to adopt consumer-grade interfaces.
At Amazon, Lewis learned a specific mode of strategic thinking: the six-page narrative document, the ruthless analysis of input metrics over output metrics, the willingness to fund initiatives that might fail in order to secure the one that transforms the business. This organizational DNA matters economically because it affects capital efficiency and strategic patience. A founder without operational experience might view trucking as a gig-economy play, treating drivers as interchangeable labor units to be optimized like ride-share passengers. Lewis, whose grandfather built a distribution business from a Seattle garage, understands that these are small business owners with capital at risk, heir trucks represent mortgages, their routes represent relationships built over decades. This comprehension shapes product decisions: the payment terms, the flexibility of load acceptance, the respect for driver autonomy that distinguishes Convoy from more extractive models.
The economic value of this leadership style manifests in organizational design. Convoy has hired from Amazon and Google not merely for technical talent but for operational discipline: he ability to build systems that function under stress, to analyze logistics networks as dynamic organisms rather than static graphs. In a business where a software glitch can strand a driver without fuel money or cause a shipper to miss a production deadline, operational excellence is not a soft cultural factor but a direct determinant of retention and margin. The cost of fixing a failed load, customer service calls, alternative sourcing, reputational damage, exceeds the cost of preventing it through better matching algorithms. Lewis’s background creates organizational incentives to invest in these prevention mechanisms, even when growth metrics might favor volume over quality.
The Fragility of Consent
Beneath the strategic questions of vertical integration and capital deployment lies a more elemental tension: the question of consent in algorithmic coordination. Convoy’s economic engine depends on reducing empty miles through optimization, yet the gains from this optimization must be shared with drivers in ways that feel equitable, or the supply side evaporates. The platform must capture enough value to justify its existence, its fees, its data collection, its insertion into transactions, while leaving enough on the table that independent operators prefer the algorithm to the broker. This is a delicate calculus. If the platform becomes too efficient at extracting surplus, matching drivers to loads with just enough margin to keep them on the road but not enough to build wealth, it becomes merely a digital version of the exploitative brokerage relationships it seeks to replace.
The fragility manifests in moments of stress. When fuel prices spike, or when a recession reduces spot rates, drivers face margin compression. In traditional brokerage, the personal relationship allows for negotiation: a broker might reduce their commission to maintain a long-term relationship, or advance payment to a reliable driver facing cash-flow problems. The algorithm has no such flexibility. It matches based on parameters, not mercy. If Convoy’s platform enforces rigid pricing during market downturns, treating drivers as fungible capacity units to be allocated at market-clearing rates regardless of individual circumstances, it destroys the trust necessary for adoption. Yet if the platform subsidizes rates or manually intervenes to help struggling drivers, it sacrifices the scalability that justifies its valuation.
This tension between standardization and adaptation defines the platform’s long-term viability. The twenty-five percent empty mile statistic represents not just inefficiency but slack: he flexibility that allows small carriers to survive shocks. A driver running empty might be returning home for a family emergency, or positioning for a high-paying contract next week, or simply refusing a load that pays below cost. Optimization algorithms tend to view such empty miles as failures to be eliminated. But eliminating them entirely requires eliminating the autonomy that makes independent trucking attractive. Convoy must optimize gently enough that drivers still feel like business owners rather than algorithmic employees, yet thoroughly enough to deliver the efficiency gains that shippers pay for.
The assumption that drivers will accept this trade-off, that they will surrender the friction of broker relationships for the efficiency of platform coordination, remains unproven at scale. The trucking industry has seen technology waves before: satellite tracking, electronic logging devices, load boards. Each promised to eliminate inefficiency; each was resisted or gamed by drivers who valued autonomy over optimization. Convoy’s network effects depend on density, but density depends on consent. If the platform miscalculates the value drivers place on independence, or if economic conditions force it to choose between driver welfare and shipper service, the network effect operates in reverse. Drivers leave, matching quality degrades, shippers return to brokers, and the data moat becomes a record of abandoned routes.
The Cold Season
The defining test for Convoy will not occur during the current expansion, when freight volumes remain high and capital remains cheap. It will come in the cold season: he next recession, when the quarter of trucks that run empty today becomes half, when shippers switch from spot markets to contracted rates with asset-based carriers, when drivers park equipment rather than haul loads at a loss. In such an environment, the platform’s value proposition inverts. Instead of offering drivers more loads than they can find through brokers, Convoy competes for scarce freight against every other digital and traditional broker. Instead of offering shippers access to scarce capacity, it offers access to desperate capacity, which is not the same thing. Desperate drivers accept any rate, service quality degrades, and the matching algorithms optimize for survival rather than efficiency.
Dan Lewis’s memory of 2001 and 2008 suggests he understands this cyclicality, yet understanding does not guarantee survival. The capital structure of a venture-backed startup, with preference stacks, growth covenants, and liquidation preferences, creates rigidities that the trucking cycle does not respect. A traditional broker can shrink to a phone and a desk, operating as a sole proprietor through downturns. A company with nearly a thousand employees and $668 million in funding cannot contract gracefully; its cost structure is predicated on growth or death. When freight rates fall below operating costs, Convoy must choose between subsidizing drivers to maintain network density, a cash bonfire, or allowing density to collapse, destroying the network effects that justify its existence.
The bear case for Convoy is not that the technology fails, but that the sociology of trucking resists aggregation more stubbornly than expected, and that the capital required to force aggregation destroys returns in the process. If drivers retain sufficient loyalty to traditional brokers, or if economic downturns reveal that the platform’s efficiency gains are cyclical illusions, artifacts of tight capacity rather than structural improvements, Convoy becomes merely another broker with higher fixed costs. The data moat, impressive during expansion, proves shallow when liquidity evaporates and drivers revert to the relationships that offer payment flexibility and human discretion.
Yet if Convoy navigates these tensions, aintaining driver consent through downturns, preserving capital discipline while achieving scale, resisting the temptation to vertical integrate while ensuring quality, t could transform the economics of American freight. The twenty-five percent empty mile represents billions in wasted fuel, time, and carbon emissions. A platform that genuinely reduces this waste while preserving the independence of small carriers offers a rare alignment of economic and social value. The question is whether such alignment is possible, or whether the imperatives of venture scale and algorithmic efficiency inevitably extract more value than they create, leaving behind another optimized ruin when the cycle turns.
Consider the mechanics of a freight recession. When industrial production contracts, the first cut comes from inventory in transit. Shippers cancel spot market tenders and retreat to contracted capacity with asset-based carriers who cannot easily shed obligations. The spot market, where Convoy derives its margin and its liquidity, does not merely shrink, t collapses. Spot rates can fall forty percent in a quarter, dropping below the operating cost of a small carrier. In such moments, the independent owner-operator faces a binary choice: haul at a loss to cover fixed payments on the tractor, or park the truck and default on the loan. Most choose a third, precarious option, xiting the spot market entirely and signing on as a company driver for a large fleet, surrendering the independence that defines their business model in exchange for a guaranteed paycheck and health insurance.
This exodus presents a specific death spiral for digital brokerages. When drivers park their authority and accept employee status, they vanish from the platform not temporarily but structurally. The network density that Convoy has spent hundreds of millions to achieve evaporates not because the technology failed, but because the economic substrate, the independent owner-operator, has temporarily ceased to exist as a market participant. The data moat becomes a historical record of routes no longer driven by carriers no longer operating. When the market recovers eighteen months later, those drivers do not automatically return to the platform; they have settled into employment, sold their equipment, or aged out of the industry. Rebuilding density requires acquiring a new generation of carriers who have no memory of the previous platform and no reason to trust it.
The cash flow dynamics compound this fragility. Trucking operates on a brutal liquidity asymmetry: drivers demand payment within twenty-four hours, fuel must be purchased, tires replaced, loans serviced, while shippers universally pay net thirty or net sixty. Traditional brokers bridge this gap through factoring, advancing cash to drivers at a discount, or through deep balance sheets built over decades of accumulated retained earnings. Convoy’s venture capital, designed for growth rather than working capital intensity, faces a sudden liquidity trap during downturns. If the platform maintains its policy of fast payment to drivers while shippers extend payment terms to conserve cash, the burn rate accelerates precisely when revenue contracts. A marketplace model that requires factoring infrastructure to survive becomes, in effect, a lender to a failing industry, absorbing credit risk from bankrupt shippers and desperate drivers without the regulatory framework or loss reserves of a financial institution.
The comparison to Uber Freight, which entered the market with deeper pockets and superior technology but has struggled to achieve density, illustrates the specific resistance of trucking to gig-economy logic. Ride-share drivers transport humans who value consistency and safety standards; they operate within urban geographies where density is achievable; they accept algorithmic dispatch because the alternative is unemployment or minimum wage service work. Truckers, by contrast, transport commodities where price is the primary signal; they operate across continental distances where density is inherently sparse; and they operate capital equipment that represents significant equity, making them business owners rather than labor inputs. When Uber Freight offered guaranteed rates to drivers during the 2019 downturn, it discovered that truckers accepted the loads while continuing to search for better rates on other platforms, arbitraging the guarantee while maintaining multi-homing behavior. The platform could not enforce exclusivity without becoming an employer, yet could not achieve liquidity without it.
This multi-homing: he ease with which a driver can accept a Convoy load on Monday and a traditional broker’s load on Tuesday while checking a dozen load boards in between, revents the lock-in that sustains network effects in consumer platforms. The switching cost for a driver is not the abandonment of accumulated data, but merely the download of another application. During expansion, this frictionless competition drives efficiency; during contraction, it drives commoditization. When every load is contested by desperate carriers, the platform cannot maintain take rates without losing supply to competitors willing to operate at breakeven or loss. The margin compression does not discriminate between digital and analog brokers; it simply reveals that software overhead, however lean, remains overhead in a market where the lowest-cost coordinator wins.
The specific precipice, then, is this: Convoy has built a coordination engine designed for surplus, optimizing the allocation of scarce capacity across abundant demand. It has not yet proven it can survive the inverse: world of abundant capacity and scarce demand where the optimization problem becomes not matching but solvency. In such a world, the personal broker with a decade of relationship equity can call in favors, extend informal credit, and weather the storm through social reciprocity. The algorithm can only match what exists; it cannot generate trust ex nihilo when the margin for error disappears. The twenty-five percent empty mile, that inefficiency which Convoy seeks to eliminate, serves during contractions as a pressure valve: space where the industry shrinks without exploding. Remove that slack through perfect optimization, and the system becomes brittle, breaking precisely when resilience matters most.