By Luis Lopez, AI transportation consultant, CEO of Go Hub.io Holdings Corp and subsidiaries, and host of the Freight Guru Podcast
Most freight brokers prospect the same way they always have: buy a list, load it into an email tool, and send the same message to thousands of people. The result is low reply rates, annoyed recipients, and a sales team that burns out chasing contacts who never had freight to move.
AI can change that, but only if it is used to find better signals and write better outreach, not to send more noise. As an AI transportation consultant, the first thing I look at in a broker’s sales process is whether they can explain why a given shipper is on their list right now. If the answer is “they were in the database,” the problem is not volume. It is targeting.
Start with public signals of freight need
Shippers leave clues about their freight. AI tools are good at collecting and sorting those clues at scale. Useful categories include:
- Import records in general. Some trade data services compile information from ocean shipping manifests. A company that imports containers regularly through a nearby port may need drayage, transloading, or outbound truckload capacity. Treat this data as directional, since company names and details are often incomplete or inconsistent.
- Job postings. A company hiring logistics coordinators, transportation managers, or warehouse staff in a new city is often expanding or reorganizing its freight.
- Facility openings. New distribution centers, plants, and stores generate inbound and outbound freight. Local news, permits, and company announcements can surface these early.
- Seasonal patterns. Many industries have predictable peaks. Reaching a lawn and garden supplier or a holiday goods importer before their busy season is far more useful than calling in the middle of it. Our guide to seasonal freight patterns can help you map timing.
- Product launches and new markets. Expansion into new regions usually means new lanes.
No single signal proves a shipper needs capacity. Combined, they show where a conversation is likely to be welcome.
Lead scoring: ranking who to call first
Lead scoring takes those signals and assigns each prospect a priority. A simple model might weigh how closely the shipper’s freight matches your strengths, how recent and strong the signals are, and whether you have served similar shippers well before.
Machine learning models can go further by learning from your own history: which kinds of prospects turned into accounts, which stalled, and which became problem customers. The key is to train on outcomes that matter, such as loads moved and payment behavior, not just meetings booked. A shipper that books plenty of meetings but never tenders freight is not a good lead.
Credit risk belongs in the score as well. A prospect with lots of freight and poor payment history can hurt you. See our article on how AI helps brokers assess shipper credit risk for more.
Enrichment: filling in the picture
Enrichment means adding details to a basic company record: industry, locations, size, likely decision makers, and the equipment types their freight probably needs. AI tools can pull this from multiple sources and summarize it for a salesperson in a few lines.
Two cautions apply. First, enrichment data is often stale or wrong. Contacts change jobs, and facilities close. Second, more data is not always better. A rep needs three relevant facts, not a forty-field profile.
Personalizing outreach without becoming spam
Generative AI can draft emails that reference a shipper’s actual situation: a new facility, a seasonal peak, or a lane you already run. Done well, this makes outreach more relevant. Done badly, it produces thousands of messages that are technically personalized but obviously automated.
Some practical rules:
- Lead with something specific and true about the shipper, not a compliment.
- Offer something useful, such as lane insight or a capacity option, not a generic list of services.
- Keep messages short, and have a person review drafts before they go out, at least until you trust the process.
- Check every factual claim the AI adds. Language models can invent details that sound real. Our piece on catching AI hallucinations in freight explains how.
- Limit volume. Fewer, better messages protect your sender reputation and your brand.
Know the basic email rules
In the United States, the CAN-SPAM Act sets ground rules for commercial email. At a general level, commercial messages must not use deceptive headers or subject lines, must include a working way to opt out, and must include the sender’s physical postal address. Opt-out requests must be honored. AI does not change these obligations; it just makes it easier to break them at scale. Make sure your tools enforce them automatically, and talk to counsel about any other rules that apply to your outreach.
Data quality decides the results
A prospecting system is only as good as its data. Duplicate company records, outdated contacts, and mismatched names between data sources will send reps in circles. Before adding AI, clean up your CRM: merge duplicates, standardize company names and addresses, and record why deals were won or lost. That history is what makes scoring models useful later.
Set a simple feedback loop. When a rep finds a bad record or a wrong signal, it should be easy to flag and fix. Over time, the system gets better instead of drifting.
The human relationship still closes the deal
Shippers do not hand freight to a broker because an email was well written. They give freight to people who answer the phone, solve problems, and tell the truth when a load is running late. AI can help a rep reach the right shipper at the right time with a relevant message. It cannot build trust on its own.
The best use of AI in sales is to remove the research grind, so reps spend more of their day in real conversations and less of it digging through spreadsheets.
What not to automate
- Pricing commitments. Quotes should come from your pricing process, not from a sales email generator.
- Capacity promises. Never let a tool promise trucks or service levels your operations team has not confirmed.
- Follow-up on sensitive issues. Service failures, claims, and billing disputes need a person.
- Endless sequences. If a prospect has not replied after a reasonable number of messages, stop.
- Anything you would not say face to face. If a message would embarrass you in person, it should not go out under your name.
A practical starting point
Pick one or two industries where you already perform well. Define the signals that suggest a shipper in that space needs capacity. Build a short, scored list, enrich it lightly, and have reps send personal, reviewed messages. Track replies, meetings, and, most importantly, loads moved. Then adjust. That small, disciplined loop will beat a mass-email blast every time.
For more conversations on freight and AI, subscribe to the Freight Guru Podcast.
About the author: Luis Lopez is a Miami-based AI transportation consultant and logistics entrepreneur, the CEO of Go Hub.io Holdings Corp and subsidiaries, and host of the Freight Guru Podcast.
