4 signals your routing guide is about to break 

This is part 2 of a two-part series on how AI is reshaping freight procurement. Read Part 1: From Event-Based to Proactive →

Your routing guide looks fine. The data says otherwise. 

Most procurement teams do not discover routing guide problems at the moment they start. They discover them weeks or months later, when the damage has already compounded, in budget overruns, service failures, or spot market exposure that nobody planned for. 

The reason is simple. Traditional freight procurement was not built to detect problems in real time. It was built to negotiate contracts on a schedule and then assume those contracts hold until the next cycle. Technology (or lack thereof) forced a choice.  

In a tightening capacity market, that assumption is increasingly dangerous. When fewer carriers are available, spot rates climb. When spot rates climb, carriers reject contracted loads to chase better margins in the open market. Tender rejection rates spike. Routing guides that looked solid six months ago begin to fracture. 

The signals are usually there before the routing guide fully breaks. The question is whether your systems are surfacing them, or whether you are finding out after the fact. 

Here are the four signals that project44’s AI Freight Procurement Agent acts upon, what each one means for your network, and what to do about them. 

Signal 1: Deteriorating carrier performance 

What it looks like. A carrier that was performing well when awarded primary status has seen its on-time delivery decline over the past several weeks. Service levels are eroding. But because performance reviews happen quarterly, if they happen at all, nobody has connected the trend back to a procurement decision. 

Why it matters. In a tightening market, carrier performance issues are often the first sign that your routing guide is under stress. When carriers can get better rates in the spot market, their commitment to contracted loads weakens. Acceptance rates drop. Transit times stretch. Exceptions increase. These are not random service failures. They are rational market behavior, and they mean your procurement assumptions are no longer holding. 

Carrier performance is also highly lane-specific. A carrier that performs well on one corridor may be underperforming on another. Network-level OTIF metrics mask these lane-level problems, and most legacy systems do not surface them until the damage is visible in operational data. 

What the Agent does. The Agent evaluates carrier performance by lane using verified visibility data, not carrier-reported tracking. It monitors on-time delivery, tender acceptance, and tracking compliance as ongoing signals, not as periodic scorecards. When performance trends deteriorate beyond your configured thresholds, the Agent identifies the lane, surfaces the issue, and can initiate sourcing actions to add capacity or renegotiate terms. 

It also maintains a feedback loop. If your team consistently selects a specific type of carrier on certain lanes, for example, favoring a regional carrier over a national one , the Agent learns that preference and adjusts its scoring accordingly over time. 

Signal 2: Lanes running above market rate 

What it looks like. A lane that was competitively priced at your last RFP is now running well above current market benchmarks. The rate felt right when it was negotiated. The market moved. Nobody flagged it because the contract was not up for review. 

Why it matters. This is the most common source of hidden freight spend. When contracted rates drift above market, you overpay on every shipment without realizing it. Multiply that across dozens or hundreds of lanes, and the gap between what you are paying and what you should be paying becomes significant, fast. 

The challenge is that spotting above-market lanes requires comparing your contracted rates against current market conditions in real time. Most transportation management systems do not do this. They store the contract. They do not evaluate whether the contract still makes sense. 

What the Agent does. The AI Freight Procurement Agent benchmarks your contracted rates against live market data across your lane portfolio. When a lane crosses a configurable threshold above market, the Agent flags it and calculates the potential savings based on your volume on that lane. It does not just surface the problem. It launches a targeted mini-bid to re-engage the carrier market and negotiate rates that reflect current conditions. 

Information flows from project44’s logistics data graph, over 1.5 billion shipments annually across 259,000+ carriers. This is observed market data from actual execution, not rate estimates or carrier-reported figures. Coupled with external market data, the Agent operates from a complete, verifiable picture of real-world activity. 

Signal 3: Uncovered lanes with rising spot exposure 

What it looks like. A lane does not have an active contract. Maybe it never did. Maybe the volume was not high enough to justify contracting it during the last RFP. But over the past several months, volume has increased. Every shipment on that lane goes to the spot market. And as spot rates climb with tightening capacity, the cost of leaving that lane uncovered grows with every load. 

Why it matters. Spot market exposure is one of the clearest indicators of procurement gaps. Some spot exposure is expected and even strategic. But uncontrolled spot exposure, where shippers are routinely reverting to spot pricing on high-volume lanes because they lack contracted coverage, is a direct hit to transportation costs. 

In the current market, this signal is especially important. Tender rejection rates above 14% mean more loads are falling out of the routing guide and into the spot market. Each rejected load becomes an unplanned spot purchase at whatever rate the market will bear that day. 

What the Agent does. The Agent identifies lanes with elevated spot exposure by analyzing your booking patterns and comparing contracted versus spot utilization. When it identifies a lane where spot activity exceeds a configurable threshold and volume justifies a contract, it evaluates carriers and recommends or launches a mini-bid to secure contracted coverage. 

This is one of the most powerful use cases for carrier discovery. The Agent does not just renegotiate with carriers you already know. It identifies carriers from across the network that run your lanes well, have the right equipment, and carry the right insurance , carriers outside your current “bubble” that you would not find through manual processes. 

Signal 4: Expiring contracts 

What it looks like. A contract is approaching expiration. The carrier is performing well. The rate is competitive. But the renewal date is buried in a spreadsheet alongside hundreds of other contracts, and nobody has initiated a conversation with the carrier. 

Then the contract lapses. The lane reverts to spot pricing. You pay more for the same capacity from the same carrier because nobody got around to extending the agreement. 

Why it matters. This is the simplest problem on this list and one of the most expensive. Contract lapses are not a procurement strategy failure. They are an execution gap, a consequence of managing renewal timelines manually across a large network. 

In a market where capacity is tightening and carrier availability is shrinking, losing a performing carrier on a key lane because of an administrative miss has real consequences. Rebooking that lane at spot rates in a tight market costs significantly more than renewing the existing contract would have. 

What the Agent does. The Agent monitors contract expiration dates across your portfolio and initiates renewal outreach to carriers based on a configurable window, for example, 30, 60, or 90 days before expiration. If the carrier is performing well and the rate is competitive, the Agent can extend the contract proactively. If conditions have changed, it can launch a mini-bid to evaluate alternatives. 

Every contract the Agent negotiates is presented for human approval before it goes live. Nothing is auto-accepted. Your team reviews the negotiated terms, sees the projected impact on cost and performance, and confirms or adjusts. 

These signals can be the most costly 

Each of these four signals tells you something specific about a lane in your network. But they are not isolated events. They are symptoms of the same structural dynamic: static procurement processes struggling to keep up with a market that moves faster than manual intervention allows. 

The capacity cycle reinforces all of them. Capacity tightens. Spot rates increase. Carriers reject contracted loads to capture higher spot margins. Tender rejection rates climb. Lanes drift above market. Performance erodes. Contracts lapse. Spot exposure grows. And the cycle accelerates. 

Addressing these signals one at a time, reactively, after they surface in a quarterly review, is how most shippers have always operated. It was good enough when the market was more predictable. 

It is not good enough now. 

What makes project44’s AI Freight Procurement Agent different from analytics dashboards or benchmarking reports is that it closes the loop. It does not just surface signals. It acts on them,launching mini-bids, engaging carriers, managing negotiations, and presenting ready-to-approve contracts,within the guardrails your team defines. 

And it does this whether you are using project44’s Intelligent TMS as your primary system or running the Agent as a standalone module alongside your existing TMS. The modular design means you do not need to re-platform your entire transportation management stack to start detecting and responding to these signals. 

The shippers who gain structural advantage in this market will not be the ones with the best annual RFP. They will be the ones whose procurement systems detect shifts as they happen and act before the routing guide breaks. 

---