AI in freight forwarding: where to start, and where not to 

The best AI wins in freight are boring. And that is the point. 

For the last two years, the conversation about AI in logistics has been full of big promises: self-running supply chains, robots in every warehouse, decisions made without people. Meanwhile, the forwarders seeing real results are doing something much less exciting. They are using AI to take the slow, repetitive work off their teams’ desks. 

If you are wondering where to start, here is our view. 

Start where the work is repetitive and the rules are clear 

AI works best on tasks that happen hundreds of times a week, follow a pattern and currently eat up skilled people’s time. In freight forwarding, three areas stand out. 

1. Documents 

Commercial invoices, packing lists, bills of lading, customs paperwork. Every shipment comes with a stack of documents, and someone has to read them, check them and key the details into your system. AI can now read these documents, pull out the key information and flag anything that does not match. Your team checks the exceptions instead of typing every line. 

2. Quotes 

Quoting is where many forwarders win or lose business. It is also slow: finding the right rates, adding surcharges, checking margins and writing it all up. AI can assemble a first draft of a quote from the rates you already have, so your sales team spends their time on the customer, not the spreadsheet. Faster quotes win more business. 

3. Exceptions 

Delays, missed cut-offs, documents that have not arrived, shipments that have not moved. Today, many teams find these problems when the customer calls to ask. AI can watch your shipments and raise a flag early, so your team can act before the customer notices. 

None of these replace your people. They give your people back time for the work that needs judgement, experience and relationships. 

Where not to start 

Just as important is knowing where AI is not the right first step. 

Do not start with a big-bang project. A programme that tries to transform everything at once is expensive, slow and hard to prove. Start with one problem, show results and build from there. 

Do not start without clean data. AI learns from your data. If your customer records are duplicated, your codes are inconsistent and your rates are out of date, AI will make the same mistakes faster. Fixing the basics first is never wasted. 

Do not start with the most complex decisions. Pricing strategy, carrier negotiations and customer relationships rely on experience and context. AI can support these with better information, but it should not be the first place you hand over control. 

Do not start because a competitor did. The right first project is the one that solves your biggest problem, not the one that makes the best headline. 

A simple way to begin 

If you are not sure where to start, ask your team one question: what task takes up the most time every day, and adds the least value? The answer is usually a good first candidate. 

Then keep it small. Pick one team or one lane, agree how you will measure success, run it for a few weeks and look at the results honestly. If it works, scale it. If it does not, you have learned something useful at a low cost. 

The bottom line 

AI in freight forwarding does not need to be dramatic to be valuable. The quiet wins in documents, quotes and exceptions free up time, reduce errors and make your team’s day easier. That is a strong place to start. 

At ZBC, we help logistics businesses find the right first step, whatever systems they use today. 

Where has AI saved your team time so far? Talk to us.