What Twenty Years of Freight Technology Actually Changed

There's a moment familiar to anyone who has worked a freight desk. A tracking screen shows a truck stopped on a shoulder outside Kingsville. It has been stopped for two hours. The dot is accurate, the timestamp is current, the geofence is working exactly as designed. And it tells you nothing you can use.

Is the driver asleep? Broken down? Waiting on a load lock? Is this a delay you can absorb or a delay you need to call the customer about right now?

The technology did its job. It found the problem. What it can't do is tell you what the problem means, and it certainly can't fix it.

That gap — between visibility and resolution — is the most useful lens for understanding what two decades of freight technology has and hasn't accomplished. It's worth walking through properly, because the industry talks about this badly. Vendors oversell. Skeptics dismiss. The reality is more interesting than either.

Three waves, briefly:

Freight technology didn't arrive all at once. It came in roughly three waves, and each solved a genuinely different problem.

The first wave was matching. Before load boards went online in the 1990s, finding a truck for a load meant a Rolodex, a phone, and a lot of luck. The board turned a local search into a national one. This was an enormous change and it's easy to forget how much it mattered. It also had a side effect nobody planned: it made freight pricing transparent enough to commoditize, which compressed margins across the whole brokerage industry and never uncompressed them.

The second wave was measurement. The ELD mandate, phased in starting December 2017, is the cleanest dividing line in modern trucking. Electronic logging replaced paper logbooks, which meant hours of service became enforceable in a way they had never been. Capacity that had quietly existed on paper — the extra two hours a driver could stretch — disappeared from the market almost overnight. Rates moved. Small carriers who had operated in the gray exited. And as a byproduct, every truck now carried a device that knew where it was, which made real-time visibility technically possible for the first time at scale.

The third wave was integration. Visibility platforms, API connections between shipper TMS and broker systems, automated appointment scheduling, dynamic pricing engines trained on historical lane data. This is the wave the industry is still in, and it's the one that gets the most marketing attention.

Each wave delivered something real. None of them delivered what the pitch decks promised.

The visibility paradox:

Here's the strange thing about the third wave: it made shippers less patient without making freight more reliable.

Before real-time tracking, a shipper found out about a late truck when it didn't show up. That was bad, but it set expectations accordingly. You built buffer into your schedule because you had no choice.

Now a shipper watches the truck the entire way. They see it stop. They see it sit. They see the ETA slip by twenty minutes, then forty. The information arrives continuously, which means the anxiety does too — and critically, the information arrives without any accompanying explanation.

The result is a strange inversion. Visibility platforms were sold as a way to reduce the number of "where's my truck" calls. In practice, many shippers report making more contact with their brokers, not less, because now they can see something happening and want to know what it means.

This isn't an argument against visibility. It's an argument that visibility is an input, not an outcome. Data creates questions. Somebody still has to answer them.

What happened to the digital freight brokerages:

The most instructive story in freight technology is the rise and partial collapse of the venture-funded digital brokerage.

The thesis was clean and genuinely compelling: brokerage is a matching problem, matching problems are software problems, and a company that automated matching could operate at a fraction of the cost of a traditional broker with its floors of salespeople. Billions of dollars went into this idea. Uber Freight, Convoy, Transfix, and a long list of others.

Some of it worked. Instant quoting on standard dry van lanes is now table stakes, and that's directly downstream of competitive pressure from digital entrants. App-based load booking made life meaningfully better for small carriers. These were real contributions.

But the model ran into a wall, and Convoy shutting down in October 2023 after raising well over a billion dollars made the wall hard to ignore. The wall was this: the loads that automate well are the loads with the thinnest margins, and the loads with real margin are exactly the ones that resist automation.

A dry van going from a large DC to another large DC, both with standing appointments and flexible windows, is a software problem. Software should handle it, and increasingly does.

A flatbed with an oversize load going to a job site with no dock, a crane scheduled for Thursday morning, and a permit that expires Friday is not a software problem. It's a sequence of judgment calls, phone calls, and contingency plans. And it prices accordingly.

Automating the easy half of a business while the hard half still requires people means you've built a lower-margin version of the same company. That's roughly what happened.

Where technology hits a hard limit: the border:

Cross-border freight between the U.S. and Mexico is the clearest case study in what software can and can't do, because the constraints are physical and legal rather than informational.

A truckload moving north from Monterrey typically involves three carriers, not one. A Mexican carrier brings the freight to the border. A transfer carrier — a transfer or drayage operator working the short shuttle — moves it across. A U.S. carrier picks it up on the other side and runs it to destination.

Each handoff is a place where a shipment can stall, and each handoff involves a different company with different systems, different visibility tools, and often a different language. There is no single platform that sees all three. Even where each leg has tracking, the shipment as a whole doesn't — it appears, disappears, and reappears.

Layer on the documentation. Mexico's Carta Porte complement, the supplement to the CFDI electronic invoice required for domestic freight movements, has to be correct and correctly linked. Customs entries have to be filed and cleared. A U.S.-bound load also needs its ACE manifest transmitted in advance.

And here's the part software doesn't solve: when the paperwork has a defect, the truck stops. Not "the system flags a warning" — the truck physically stops at the crossing, and it stays there until a human being finds the error, gets the corrected document issued by whoever issued the original, and gets it into the right hands.

Every hour of that is real money. Cross-border detention is expensive, border-area parking is limited, and cargo theft risk in staging areas is not theoretical.

The most valuable intervention in that entire chain happens before the truck ever leaves the shipper: someone reviewing the documentation package and catching the defect while it's still cheap to fix. That's a checklist executed by a person who knows what to look for. It is not a feature.

What's actually improving now:

None of this means the technology story is over. Several things are genuinely getting better, and they're worth knowing about.

Appointment scheduling is finally being solved. Dock scheduling has been an absurdly manual process — email chains, phone trees, spreadsheets — and integrated scheduling tools are meaningfully reducing detention at facilities that adopt them. This is unglamorous and it's the single biggest lever on driver productivity available right now.

Predictive ETAs are getting good. Not "the truck is here" but "based on this route, this weather, this driver's remaining hours, and historical dwell at this receiver, here's when it will actually arrive." That's genuinely actionable in a way a live dot isn't, because it converts data into a decision.

Fraud detection is improving because it had to. Double-brokering and identity theft in freight got dramatically worse in the early 2020s — bad actors registering lookalike carrier authorities, taking possession of loads, and disappearing. Carrier vetting tools have gotten sharper in response, cross-referencing authority history, insurance changes, phone and email provenance. This is one area where the software genuinely does something a person can't do as fast or as thoroughly.

Document automation is quietly the biggest one. Extracting data from rate confirmations, BOLs, and PODs without manual entry removes an enormous amount of low-value work and a corresponding amount of human error. It doesn't market well. It matters a lot.

Notice the pattern in all four. The technology that's actually delivering isn't the technology that replaces judgment. It's the technology that removes the tasks surrounding judgment, so the people doing the work have more room to do it.

What this means if you're the one tendering freight:

A few practical conclusions.

Treat visibility as a floor, not a differentiator. If a broker can't give you tracking in 2026, that's disqualifying. But every serious competitor has it, so it tells you nothing about which one to choose.

Ask what happens after the alert. The useful question isn't "do you have real-time tracking." It's "when the tracking shows a problem, who sees it, what do they do, and when do I hear about it." The answers vary enormously and the difference is entirely operational.

Match the freight to the model. If you're moving repeatable dry van lanes between fixed facilities, an automated platform will probably serve you well and price competitively. If you're moving anything with a constraint — oversize, a job site, a border crossing, a tight receiving window — you want a person whose name you know, and you should expect to pay for the difference.

Look at how they handle exceptions, not how they handle normal. Every broker looks the same when the load goes fine. Ask a prospective partner to walk you through a specific load that went wrong, what they did, and what it cost them. The ones with a good answer will have it ready.

The honest version

Technology in freight has done a lot. It found the trucks, measured the hours, and made the paperwork move faster. Anyone romanticizing the days of the paper logbook and the Rolodex wasn't there.

What it hasn't done — what it shows no particular sign of doing — is remove the need for someone to make a decision when a load goes sideways at 4pm on a Friday. Freight is a physical business conducted across thousands of independent operators, and the failure modes are stubbornly analog. Trucks break. Drivers get sick. Ports congest.

Documents have typos. Receivers close early without telling anyone.

The best operators in this business use every tool available and understand exactly what those tools are for. They're not skeptics and they're not evangelists. They just know that the dot on the screen is where the work starts, not where it ends. EQ Logistics is a truckload freight brokerage covering domestic and cross-border lanes, with operations based in Monterrey.