Why I Stopped Trusting DAT, FreightWaves, and Truckstop to Tell Me What My Freight Is Worth
By Millisa Nwokolo, Founder of La Crown Inc. — April 8, 2026
A carrier can be in your sweet spot every single load — and still be costing you money. The SaaS platforms can't show you that. But your own data can.
Something a freight broker named Seth Bland posted on LinkedIn stopped me mid-scroll this week. He shared that his system tracks profit and loss by carrier — and what it revealed was something every broker in this industry has felt but rarely quantifies. He had a carrier who was always in his sweet spot. The right lane. The right equipment. The right response time. On paper, this was a model relationship. But his data told a different story: he was losing money on that carrier every single time.
Let that land for a second. A carrier who looks perfect on the surface — and is quietly bleeding your margin dry underneath. And here's the part that doesn't get said out loud enough: that bad data is now sitting inside the rate engines of the platforms you're paying to tell you what a lane is worth.
DAT. FreightWaves Sonar. Truckstop. These are tools I've used. Tools a lot of us have used. And they have real value for certain things. But when it comes to telling you what your lane is actually worth — accounting for your accessorials, your carrier relationships, your detention patterns, your fuel exposure — they cannot do it. Because they don't have your data. They have everyone's data, averaged together, losses included.
The Problem with Borrowed Data
Here's how the SaaS rate engines work. They aggregate transaction data from carriers and brokers across the country. They compile it, normalize it, and present you with a rate range for a given lane. It sounds scientific. It looks authoritative. And for a broad market view, it's useful directional information.
But that aggregated data includes every loss someone else ate to cover a load. Every detention charge that never got billed. Every fuel surcharge that was underquoted because the broker used last week's diesel price instead of today's. Every accessorial — tarping fee, layover, liftgate, inside delivery, lumper — that got absorbed into the linehaul rate because nobody wanted to have the conversation with the shipper. All of it. Averaged together. Presented to you as "market rate."
Seth's carrier was in his sweet spot. But what nobody sees in a sweet spot is whether the actual all-in cost — carrier rate plus fuel exposure plus any accessorials billed or absorbed — is profitable. The rate platforms can't separate that for you. They never could.
The market rate everyone is quoting from includes everyone else's losses. You're not benchmarking against success. You're benchmarking against the average of everyone's mistakes.
According to ARK TMS's April 2026 analysis, the spot-to-contract spread has narrowed to roughly $0.11 per mile nationally — meaning the margin buffer that used to cover mispriced freight is nearly gone. A single re-cover, a tender rejection, a deadhead-heavy carrier choice — any one of these can wipe out the expected contribution margin on a load that "looked fine" at booking. In a tight market, borrowing someone else's rate data isn't just imprecise. It's dangerous.
The Accessorial Nobody Is Accounting For
Let me walk you through what a "rate" actually needs to contain before it becomes a real number. Because this is where the platforms fundamentally break down — and where your own data becomes worth more than any subscription you're paying for.
The Diesel Problem Nobody Is Solving Fast Enough
I want to spend a moment on fuel because this is where the gap between "market rate" and "your actual cost" is widest right now. The EIA publishes on-highway diesel prices every Monday. It is free. It is public. It is the most accurate, current, government-sourced fuel data available in the country. And most brokers are not using it in their quoting workflow in any automated way.
| Region | Diesel (3/23/26) | Change vs. 1 Year Ago |
|---|---|---|
| U.S. National Average | $5.375 | ▲ $1.808 |
| East Coast (PADD1) | $5.480 | ▲ $1.805 |
| Midwest (PADD2) | $5.160 | ▲ $1.669 |
| Gulf Coast (PADD3) | $5.134 | ▲ $1.869 |
| Rocky Mountain (PADD4) | $5.174 | ▲ $1.759 |
| West Coast (PADD5) | $6.310 | ▲ $2.081 |
| California | $6.870 | ▲ $2.094 |
Look at those numbers. West Coast diesel is at $6.31. California is at $6.87. If you are quoting a lane that originates or terminates on the West Coast and you're using a fuel surcharge table that hasn't been refreshed to this week's EIA data, you are leaving money on the table on every single load. And you're doing it while trusting a platform's "market rate" that may have been calculated when fuel was a dollar cheaper.
The solution I built pulls the EIA data directly into my quoting skill. Every quote I generate is automatically indexed to the current week's diesel average for the relevant PADD region. Not last month. Not last week. This week. That's the difference between a rate that's accurate and a rate that just looks accurate.
The Capacity Wall That's Coming
There's another reason the platforms' historical rate data is becoming less reliable — and it has nothing to do with how the platforms work. It has to do with how the carrier market is changing beneath all of us.
According to IFA Commercial Factor's 2026 analysis of FMCSA data, an estimated 88,000 trucking authorities were revoked in 2023 alone — followed by a net contraction of nearly 10,000 additional motor carriers in just the first half of 2024. That capacity didn't come back. And now, as Truckstop.com and Bloomberg Intelligence confirmed in their February 2026 survey, federal enforcement against noncompliant drivers and carriers is actively pushing more capacity out of the market.
What that means for rate accuracy: the carriers who were running loads at distressed rates during the freight recession — the ones whose transactions built out the historical rate data in these platforms — are largely gone. The rates they were accepting are gone with them. But their data is still in the engine. Still being averaged. Still being presented as "market."
Trucking authorities revoked in 2023 alone — their distressed rates are still in the platforms' historical data
FreightWaves noted in their 2026 broker strategy piece that the spread between what brokers can charge and what they must pay carriers is narrowing as survivors demand rates that actually cover their costs. The platforms will catch up to this eventually. But eventually doesn't help you today when you're quoting a load on a lane where carrier costs moved last week.
How We Built Our Own Engine
This is the part I want to be real with you about — because this didn't happen overnight, and it didn't happen in a vacuum. I built these rate skills over months of working on real quotes, for real customers, on lanes I actually run. Not hypothetical lanes. Not demo data. My book.
The foundation of the system is Claude Code and a set of SKILL.md files — precise instructions that tell the AI exactly how to handle a specific customer, a specific lane, a specific equipment type. Each skill is built around what I actually know:
mileage.py in the freight-calculator skill holds coordinates for 100 major freight hubs and calculates road miles using a geodesic calculation with a 1.22 road factor — the standard freight approximation. For anything not in the database, it geocodes automatically via OpenStreetMap. The miles drop straight into the quote. No tab-switching. No estimating.What the Platforms Can't Build For You
I want to be fair here. DAT, FreightWaves Sonar, Truckstop — these are legitimate tools. For broad market intelligence, for benchmarking at 30,000 feet, for understanding what a lane looks like nationally, they have value. I'm not telling you to cancel your subscriptions.
I'm telling you to stop letting them be your primary pricing engine for quotes you're sending to customers you know, on lanes you run regularly, with carriers whose behavior you've been watching for years. For those quotes, you have better data than they do. You just haven't organized it yet.
That's exactly what building these skills with Claude did for me. It forced me to organize my own knowledge — my lane history, my carrier patterns, my accessorial exposure, my P&L by customer — into a system that could actually use it. And the result is a rate recommendation that's not the market average. It's my average, updated in real time, with every variable that actually affects my margin already factored in.
When you look at a carrier who's always in your sweet spot — same lane, same equipment, same response time — can you tell immediately whether that carrier is profitable for you? Not whether the lane is at market rate. Whether that specific carrier, on that specific lane, accounting for all accessorials, all fuel exposure, all detention patterns, is actually making you money?
If the answer is no, you don't have a rate problem. You have a data problem. And the SaaS platforms can't fix a data problem that lives inside your business. Only your own system can.
The freight-calculator skill is live right now on GitHub inside the skills/freight-calculator folder — free to download, free to use, free to build on. It includes eia_diesel.py for live EIA diesel by PADD region, mileage.py for road mile calculation, generate_quote.py for the branded PDF, and broker_config.json where you fill in your branding once and every quote reflects your company automatically. You bring your own lane history. The skill builds the engine around it. This is how the small broker starts competing with the rate science teams the big houses are paying for. By building something they can't buy off the shelf — because it's built entirely on what you know.