Where logistics AI is paying off first: freight quoting, exceptions and the 90% of requests still sent by email
AI agents that read inboxes and pull contract and spot rates are doubling throughput per person at some forwarders. For shippers buying freight, the payoff is faster quotes and fewer surprise charges, if the context behind each lane is captured.
Article summary
- Why it matters: Freight procurement still runs on email. When a forwarder or broker can price your shipment in minutes instead of hours, you get more competitive quotes and fewer missed bookings. When the same tools reconcile invoices against contracts, you catch detention and accessorial charges that slip through today.
- What to do first: Ask your forwarders and brokers how they quote. Faster, AI-assisted quoting should mean more competitive responses. Include response time in your provider scorecards.
- Target accessorials first. If you are buying AI for freight audit, start where leakage is measurable: detention, demurrage and charges that don’t match contract terms.
- ~90%
- of freight quote requests still arrive by email
- 2x
- files handled per person reported by early users
- 12–18 months
- before savings from fewer detention and demurrage charges show up
On this page
The details
In a FreightWaves interview, Dan Bailey, co-founder and CEO of logistics AI company Nexcade, said about 90% of freight quote requests still arrive by email. His company’s agents read inboxes, classify the freight, pull shipment details and run lookups against contract rates, spot markets and overseas agent portals. The goal is for an operator to start the day with 10 to 15 quotes ready for a pricing decision instead of working through each request by hand.
Bailey said teams using the platform have doubled the number of files each person handles. Without it, each quote can mean four or five emails and several rate lookups. One customer now answers overseas agents’ requests that arrive at 3 a.m., before competitors are awake. Nexcade tracks the public AI announcements of 40 to 50 leading forwarders and logistics providers, and Bailey called their volume over the past six to nine months “overwhelming.” About 75% of current volume is air and ocean, and 25% is road freight in the U.S. and Europe.
His view on where returns come from:
- Context is the hardest part. Most of the value sits in tacit knowledge about lanes, customers and agents. Turning that into context the AI can use is the main obstacle.
- Risk reduction compounds. Reconciliation and exception handling cut demurrage, detention and unexpected charges, with the effect becoming significant over 12 to 18 months.
- Speed shows up in revenue. Win rates by response time and margin by lane quantify the commercial effect.
- Shadow AI is already happening. “If you’re not adopting it, I can guarantee you that your employees are in shadow capacity one way or another,” Bailey said, pointing to widespread unsanctioned use of general chatbots.
- Run pilots from both ends. He recommends identifying strategic workflows and building data foundations from the top, while letting willing teams run small pilots from the bottom, giving companies “three or four bites at the apple” before a major rollout.
Where AI fits in the freight buying cycle
| Stage | What AI can do | What procurement should check |
|---|---|---|
| Quoting | Read emailed requests, pull rates and return quotes | Response time and quote accuracy |
| Bids and RFPs | Clean lane data and compare carrier responses | Whether the tool normalizes accessorials and fuel |
| Tendering and booking | Match loads to carriers and confirm bookings | Tender acceptance and exceptions |
| Tracking | Flag delays and exceptions from carrier messages | How exceptions reach your team |
| Freight audit | Reconcile invoices against contract rates | Recovered overcharges and disputed charges |
This mapping is our framework. Bailey’s comments focus on quoting and reconciliation, which is where measurable results usually appear first.
A simple business case
Take a forwarder team of five handling 30 quote files each a day. If AI doubles files handled per person, as Bailey says customers have seen, the same team handles 300 files a day instead of 150, without adding staff. For a shipper, the same logic applies to freight audit: if accessorial and unplanned charges run at a few percent of a $20 million freight budget, recovering even a fraction is worth more than many software subscriptions. These figures are illustrative; build the case on your own invoice data.
How to measure AI value in freight
| Measure | What to track | Where it shows up |
|---|---|---|
| Speed | Quote turnaround time, win rate by response time | Revenue for providers; faster awards for shippers |
| Productivity | Files or shipments handled per person | Cost to serve |
| Risk | Demurrage, detention and unexpected charges | Freight spend, over 12 to 18 months |
| Accuracy | Invoice exceptions and billing disputes | AP effort and audit recoveries |
| Margin | Margin by lane | Pricing discipline |
Questions to ask your forwarders and 3PLs about AI
- Which of your workflows serving us use AI today, and which are planned?
- How has quote turnaround and invoice accuracy on our account changed since you introduced it?
- What shipment data of ours do your AI tools use, and is it used to train models for other customers?
- Who reviews AI-generated quotes and bookings before they are sent?
- Will you share the measured results in our quarterly business reviews?
What procurement teams should do
- Ask your forwarders and brokers how they quote. Faster, AI-assisted quoting should mean more competitive responses. Include response time in your provider scorecards.
- Target accessorials first. If you are buying AI for freight audit, start where leakage is measurable: detention, demurrage and charges that don’t match contract terms.
- Document lane context. Write down the routing rules, customer requirements and exceptions your team knows by heart. Any AI tool, yours or your provider’s, needs it.
- Run small pilots before big rollouts. Bailey recommends pairing a leadership-led data foundation with small team experiments that give several chances to learn before scaling.
What to watch next
- Measured results. Ask vendors for before-and-after data on quote turnaround, files per person and exception costs over 12 to 18 months.
- Data rights. As AI providers build on your shipment data, watch contract terms on ownership and model training.
- Shadow AI policies. Expect more companies to approve specific tools so staff stop pasting rates into general chatbots.
The bigger picture
Every major forwarder and broker is announcing AI capabilities, and the volume of announcements is hard to evaluate. For shippers, the useful test is simple: are quotes faster, are invoices more accurate, and are fewer charges disputed? For the procurement side of the stack, see our procure-to-pay software ranking.
Questions procurement teams are asking
How do you measure ROI on AI in logistics?
Beyond time saved, track risk reduction, such as fewer detention, demurrage and unexpected charges caught by reconciliation, and revenue effects, such as win rates by quote response time and margin by lane.
What is 'shadow AI' in procurement and logistics?
Employees using unapproved AI tools, such as general chatbots, to do work when the organization hasn't provided sanctioned ones. It creates data and accuracy risks that formal adoption can reduce.
What share of freight quotes still arrive by email?
Dan Bailey of Nexcade says about 90% of freight quote requests still arrive by email, which is why AI agents that read inboxes and stage quotes are seeing early adoption.
How are most freight quotes requested today?
About 90% of freight quote requests still arrive by email, according to Nexcade CEO Dan Bailey, which is why AI tools that read inboxes and prepare quotes are drawing interest.
Where do AI tools in freight deliver measurable returns first?
Quoting speed and invoice reconciliation. Bailey said reductions in demurrage, detention and unexpected charges become significant over 12 to 18 months.
How we reported this
This analysis draws on 1 source, listed below, and reflects figures as reported on October 14, 2026. The sections on why it matters and what procurement teams should do are Procurement Signal analysis. It is not legal or financial advice. See our editorial standards.
