A Trucking Startup Fired All the Dispatchers. Turnover Dropped 50%.

The dispatcher who learned to let go

For 19 years, Mike Torres told truck drivers where to go and when to be there. Then his employer—a 300-truck regional freight carrier in the Southwest—adopted a new dispatch system, and Torres thought he was about to lose his job.

Instead, his job title changed to "load concierge." Drivers now pick their own routes from a marketplace-style app. Torres helps design the offerings and handles the complex loads nobody wants. He will not go back to the old way.

The system comes from a mid-stage U.S. logistics SaaS company that spent the last year piloting something close to heresy in trucking: no fixed schedules, no assigned loads, no traditional dispatch hierarchy. Drivers bid on or claim runs based on pay, timing, and their own preferences. Performance data—on-time percentage, safety records, fuel efficiency—determines who gets first pick on the best loads.

In an industry where annual driver turnover routinely hits 70 to 90 percent, the pilot customer running this model for nine months saw turnover drop to the mid-30s. Empty miles decreased by 5 to 8 percentage points. Operating margin improved by 3 to 4 percent, mostly from lower recruiting costs and better asset use.

How Reddit rewrote the roadmap

The CEO stumbled into the idea by reading driver complaints on Reddit and closed Facebook groups. The recurring theme: "toxic dispatch." Drivers described being jerked around on tight windows, dealing with opaque pay structures, and feeling powerless over the one thing that mattered most—when they could get home.

So the CEO embedded with a pilot customer for a week of ride-alongs. He sat in cabs, ate at truck stops, and heard the same frustration in person. The product team built the marketplace model on top of their existing transportation management system. Then they braced for pushback.

Dispatchers were terrified. The entire structure of their authority was about to dissolve. But after the first month, something unexpected happened: their jobs got easier. Instead of fighting with drivers over bad loads, they focused on making offerings attractive and solving genuinely hard problems—the odd-hours pickup, the load with complicated paperwork, the route nobody wants but somebody has to run.

Drivers, meanwhile, started optimizing around their lives. One veteran stacks runs that get him home every Friday. Another prefers overnight hauls and takes only those. A third bids aggressively on high-margin loads during the week, then takes long weekends. The carrier gets better utilization because drivers actually want the runs they take.

The larger pattern

This is not an isolated experiment. Internal marketplaces and self-assignment models are quietly spreading through B2B operations—call centers, warehouse labor management, field service scheduling. The justification is always efficiency: better asset use, lower idle time, optimized routing. But workers experience it as something closer to autonomy.

What makes the trucking case unusual is the industry itself. Long-haul freight is one of the most rigid, top-down sectors in the economy. Drivers are told where to go, when to leave, and how long they have to get there. The burnout rate reflects it. Carriers spend enormous sums on recruiting and training, only to watch new hires quit within months.

The marketplace model does not fix everything. Some loads still go unclaimed and need manual assignment. Performance data can feel invasive. There are fairness questions about who gets access to the best runs and how the algorithm decides. But the early results suggest that giving drivers a measure of control over their own time and routes changes the fundamental calculus of whether to stay or leave.

Trust as a feature

The deeper story here is about trust. For decades, dispatch has operated on the assumption that drivers need to be managed tightly or they will make bad decisions. The marketplace model flips that: it assumes drivers, given real information and fair compensation, will make rational choices that also happen to benefit the carrier.

Torres, the dispatcher-turned-concierge, puts it plainly: "I used to think my job was to control the board. Now I realize it was to make the board worth controlling."

That shift—from command to design, from authority to infrastructure—is the part worth watching. Enterprise software is increasingly not just a system of record but a system of governance. The tool is the workplace culture. And in industries where human turnover is the largest cost and the largest pain, that is starting to matter more than the database architecture.

One regional carrier let drivers choose their own loads. Turnover dropped in half. The question is not whether other carriers will try it. The question is what took so long.