Wedge

Start with mid-sized brokerages generating 50M to 200M in revenue that handle temperature-controlled or high-value freight. This niche experiences the most acute pain from missed updates and has the budget to deploy software to protect major shipper relationships. Expand by adding automated resolution capabilities like auto-rebooking via load boards, and subsequently offer the routing engine to enterprise shippers managing inbound carrier networks.

Timing

Large language models now reliably parse unstructured data from carrier emails, SMS updates, and driver notes in real-time. Previously, rules-based logic failed to capture the nuances of freight delays hidden in conversational updates, requiring human operators to categorize every issue.

Why This ICP

Freight brokers operate on razor-thin margins and lose enterprise shippers immediately when service failures occur. They feel the financial pain of manual exception management directly on their P&L, making them highly motivated buyers compared to asset-based carriers.

Size Of Prize

There are roughly 17,000 registered freight brokerages in the US. At an average annual spend of $60,000 per brokerage on exception management software and offshore tracking labor, the total addressable market is approximately $1.02B.

Gap Narrative

Freight brokers rely on manual track-and-trace teams to monitor shipments, meaning anomalies are only caught when a human reads an email or checks a portal. The industry needs a system that automatically detects deviations like weather delays, missed check-calls, or missing paperwork, instantly routing the context to the correct resolution workflow. Legacy Transportation Management Systems lack native, automated exception triage.

Defensibility

Defensibility compounds through deep integration into the core Transportation Management System and email servers of the brokerage. As the agent routes more anomalies, it maps a proprietary graph of carrier reliability and optimal resolution pathways. Once a brokerage replaces its offshore tracking labor with the agent, switching costs become prohibitively high.

Why This Thesis

An agent-based approach aligns with the asynchronous, multi-channel communication used in logistics. An AI agent ingests unstructured updates from email or SMS, classifies the anomaly, and executes the routing step autonomously, replacing the exact workflow of a human tracking representative.

Hardest Part

Suppressing false positives from noisy ELD and visibility API feeds so operations agents only receive alerts for actual SLA-threatening anomalies. If the system flags every minor traffic slowdown or momentary signal drop, brokers will immediately mute the routing channel.

Min Viable Scope

Build a routing engine solely for Full Truckload (FTL) delay anomalies originating from a single visibility provider and route the alerts to designated ops Slack channels based on account ownership. Deliberately exclude temperature anomalies, Less-Than-Truckload (LTL) tracking, predictive weather routing, and automated carrier re-booking.

Cold Start Problem

You lack the historical broker data needed to train a machine learning model on which anomalies actually result in financial loss or missed appointments. Break this by shipping a deterministic, rules-based engine for the first design partners and logging their manual triage decisions to build the proprietary training set.

Time To First Value

1–2 weeks to integrate with the broker's Transportation Management System (TMS) and baseline visibility API to start parsing live load statuses.

Data Moat Available

true

Technical Difficulty

Moderate

SAM

~$150M-250M US mid-market freight brokerages

SOM

~$10M-25M

TAM

~15k-20k North American and European freight brokerages × ~$35k-50k/yr exception routing software spend ≈ ~$500M-1B

Growth Rate

~12-18%/yr, driven by increasing supply chain volatility and margin compression forcing brokers to automate manual exception handling

Paid Comparable Spend

~$80k-120k/yr per firm on manual track-and-trace labor, check-call operators, and generic shared inbox software

Kill Thresholds

• Fewer than 30% of flagged anomalies are resolved through automated routing within 60 days
• Customer acquisition cost exceeds $15,000 after 90 days
• Fewer than 10 active dispatchers log into the platform daily by week 4
• Month-three gross customer churn exceeds 15%

Leading Metrics

• Percentage of anomalous loads resolved without human check-calls
• Average time-to-resolution for delayed or off-route shipments
• Number of manual check-calls eliminated per dispatcher per day
• Days to first automated re-route after TMS integration

What Proves Right

Mid-market freight brokerages purchase the routing software at $35,000 per year to replace their manual track-and-trace labor. Dispatchers route exceptions exclusively through the platform instead of shared inboxes or spreadsheets. Month-three net revenue retention exceeds 110% as brokers expand the system to cover additional lanes and carrier relationships.

What Proves Wrong

Brokerages refuse to disconnect from legacy visibility platforms like MacroPoint or Project44, treating the anomaly router as a redundant data silo. Dispatch teams bypass the routing logic to call drivers directly when shipments are delayed, leaving the software workflows unused. The time saved per exception fails to offset the annual license cost, leading to churn.