Freight broker virtual assistant: what to hire and what to automate

Most brokers who start looking for a virtual assistant are not really looking for a person. They are looking for the end of a specific feeling — the one where it is 8pm, the loads are covered, and you still have two hours of typing left before tomorrow starts.

A VA can end that feeling. So can AI. They cost very different amounts and they fail in very different ways, and almost nobody explains the difference before you sign up for a monthly bill.

I spent eight years brokering freight at TQL and ran a million-dollar book solo. I have seen both sides of this up close. Here is the honest comparison.

What a freight broker virtual assistant actually costs

The number depends on where you hire and whether the seat is dedicated or shared.

A broker I worked with paid $400 a week for a virtual assistant. That is $1,600 a month, or $19,200 a year. That figure lines up with what the market charges: managed offshore back-office seats in freight run roughly $1,600 to $2,600 per month for a dedicated team member. Trucking-specific VA services advertise $1,299 a month part-time and $1,999 full-time, against a fully loaded in-house transportation administrator at around $64,000 a year.

So the realistic range for the thing brokers actually buy is $1,300 to $2,600 a month, forever, going up over time.

What that money buys, and what it doesn’t

The VA the broker I knew hired was genuinely good at one specific kind of work: strict if-this-then-that tasks. Given a rule, she followed it exactly. Check the portal, copy the numbers, paste them here, flag anything that doesn’t match. That is real value and it is not nothing.

What she did not have was any knowledge of freight. She did not know a reefer from a dry van when she started. She could not look at a rate con and tell you the detention terms were unusual. Every judgment call came back to the broker, and every new situation needed a new rule written for it.

That is the actual shape of the trade. You are not buying a second broker. You are buying execution of rules you write, which means you also inherit the job of writing and maintaining the rules.

The three things a VA is better at than AI

Being honest about this matters, because “automate everything” is a sales pitch, not advice.

Anything that needs a phone call to a human who expects a human. A VA can chase a receiving clerk who won’t answer email. Have AI make that call and you will lose the customer, not save the time.

Anything that requires logging into a system that fights automation. Plenty of shipper portals are hostile to anything that isn’t a person clicking. A VA gets in. A script doesn’t.

Anything where the rule genuinely changes every time. If there is no pattern, there is nothing to automate. A person adapting on the fly beats a rule that has to be rewritten weekly.

The things AI does better, cheaper, and instantly

Reading. A VA reads a tender email in thirty seconds. AI reads four hundred of them in the time it takes to refill your coffee. Anything that is fundamentally “look at this text and pull out the parts that matter” is where the gap is widest — pulling stops, dates and reference numbers out of tender emails and rate cons is the clearest example.

Sorting and summarizing your own data. Twelve months of load history is a two-hour job for a person and a two-minute job for AI, and AI will find outliers a person skims past.

Writing the same message for the hundredth time. Carrier negotiation replies, customer updates, service failure explanations. Give it three emails you have actually sent and it will match your voice closely enough that nobody notices.

Working at 2am for free. This is the one brokers underrate. Your VA has a schedule. The work that piles up overnight sits there until their shift starts. AI does not have a shift.

The comparison that actually decides it

Put the two side by side on the terms that matter to an owner:

That last row is the real difference in risk. A VA’s mistakes look like human mistakes — a missed field, a slow reply. AI’s mistakes look like confident, well-formatted, completely made-up facts. That is why some work should never be handed to AI at all, and why the review step is the part nobody should skip.

What I would actually do

Not “don’t hire a VA.” Do this instead:

Spend one week writing down every task you would hand to an assistant. Just the list, as it happens, no editing.

Then split the list in two. Everything that is reading, sorting, summarizing, or drafting goes in the AI column. Everything that needs a phone call, a hostile portal login, or genuine improvisation goes in the human column.

Most brokers find the AI column is the longer one, and it is the column full of the tasks that were keeping them at the desk after dinner. Automate that side first. Then look at what’s left and decide whether it’s actually worth $19,200 a year, because sometimes it is — and it’s a much easier decision when the list is half as long.

The mistake is hiring first and discovering afterward that two-thirds of what you handed over was work that never needed a person.