The AI Premium: What a Freight Forwarding Business Is Worth Now

Prepared for the owners and operators of independent freight forwarding companies. Benchmarks drawn from Q2 2026 public company filings.

We had a wonderful time sponsoring this year's Logisyn M&A Club! As a companion to my position on the AI panel at the conference, we published this report on the tangible value AI is bringing to freight forwarders. 

The 1995 question

In 1995, the software wave hitting this industry was replacing typewriters. The firms that moved early spent the next decade competing above their weight class. This year's gathering opens with a harder version of that story: the current wave is bigger, it is moving faster, and the people who price forwarding businesses for a living are already sorting the industry into companies that act and companies that wait.

So the question underneath the panel — and most of the hallway conversations here — is worth writing down: where were you a year ago, what are you doing today, and what will you be able to show when this group meets again next year?

This report exists to put numbers under that question. Not sentiment about AI — the two or three figures an acquirer, a lender, or your own Monday-morning P&L review would compute about your business, and the benchmarks the largest forwarders on earth just published against them.

The quarter AI hit the P&L

For years, AI in freight lived on the strategy slide. This earnings season, it moved to the P&L.

• CHF 100-150M: KUEHNE+NAGEL'S GUIDED ANNUAL PRODUCTIVITY BENEFIT BY END OF 2027

• +60%: C.H. ROBINSON PRODUCTIVITY GAIN SINCE END OF 2022

• 32.2%: EXPEDITORS OPERATING EFFICIENCY, ABOVE ITS OWN 30% TARGET

Kuehne+Nagel guided investors to CHF 100-150 million a year in productivity benefits by the end of 2027, across more than 25,000 white-collar roles. A gross figure, their CFO was careful to say, before the cost of the AI itself.

C.H. Robinson says its AI investments are already paying for themselves. Productivity went up more than 60% since the end of 2022, and average headcount went down 10.8% year over year. This, all while its North American surface book outgrew a declining market for the thirteenth straight quarter.

DHL Global Forwarding named its benchmark out loud. Page 19 of its Q2 presentation commits to using digitalization and AI to "outperform digital-first competitors." When the third-largest forwarder on earth defines success as beating software-first operators, the argument about whether this matters is effectively over.

Expeditors posted 32.2% operating efficiency — above the 30% target it set for itself — before any benefit from its announced restructuring lands.

The fight is no longer over freight rates you don't control. It's over how much of your gross profit survives your own cost structure.

Forwarding giants care about one formula

Strip the labels and the giants keep score with the same formula: operating profit divided by gross profit. Of every dollar your team earns after paying the carrier, how many cents survive your own operation?

Source: Q2 2026 company filings.

McKinsey's 2022 read of forwarder economics put typical conversion at 20-30%. The scoreboard shows a spread from 17% to 42.4% — same industry, same quarter, more than 25 points of range. Scale is not what separates them.

Read the footnotes before comparing line by line: DSV strips special items, C.H. Robinson strips restructuring, and Expeditors strips nothing. Its number carries a $25 million restructuring charge and a $16 million property-sale gain inside it.

The metric has a twin, and the giants manage by both: shipments per person. C.H. Robinson tracks shipments per person per day. At independent scale the honest version is files per operator per month: total shipments handled, divided by operations headcount. Conversion tells you where you stand; files per operator tells you which way you're moving.

Productivity moved the metric

• THE CLEAN CASE / C.H. Robinson: Average headcount fell 10.8% year over year and NAST volume still grew about 1.5% against a market down 3.3% — the Cass Freight Index's fifteenth consecutive down quarter. Global Forwarding productivity improved more than 15% in the quarter; across both divisions it is up more than 60% since the end of 2022, which the company publicly credits to a strategy running hundreds of AI agents trained to specific jobs across the shipment lifecycle. The quarter's cleanest line: 96 cents of every new dollar of adjusted gross profit fell through to adjusted operating income. That's cost structure, not rate luck.

• THE FORWARD-LOOKING CASE / Kuehne+Nagel: The CHF 100-150 million program covers more than 25,000 white-collar roles carrying about CHF 1.7 billion in cost. The company declines to credit AI alone: in its words, AI, workflow standardization and centralization, and software development gains are "combined and inseparable." Air Logistics EBIT rose 35% on turnover up 20% — volume and mix doing real work alongside the program.

• THE PRECISION CASE / Expeditors: The 32.2% arrived before the restructuring shows up, as second-quarter headcount was essentially flat against Q1, so that ratio is what they did with the people they already had. The restructuring itself is expected to take roughly $50 million a year out of the cost base, close to 10% of total corporate overhead, landing mostly in Q3.

Productivity is the lever the giants talk about most. Rates, volume and mix are inside every one of these numbers too — and for DSV and Kuehne+Nagel, so is M&A. Nobody's quarter is one lever. But productivity is the one they all say they can repeat.

What buyers ask now

Ask anyone who has been through diligence recently and they'll tell you the first question has changed. It used to be volume: TEUs, tonnage, revenue. Now it is: show me productivity per head, and prove it's improving. Four screens, in the buyer's own order:

1. Operating leverage per operator: Files per operator per month, trended over eight quarters. A rising line is an asset. A flat line at low volume is a staffing model the buyer has to fix — and prices accordingly.

2. One system, or a Frankensystem: How many systems does one shipment touch between quote and invoice? Every swivel-chair hop is integration cost the buyer will subtract — and a data seam the buyer won't trust.

3. Data cleanliness: Can you produce gross profit per shipment, by lane, by month, this week? If the answer takes longer than the meeting, the buyer assumes the number is worse than it is.

4. Key-person risk: If two people left, would the margin leave with them? Process that lives in software survives a departure. Process that lives in someone's head is a discount.

The structural question behind all four

The panel this week takes up a question this room has argued about for thirty years: does technology consolidate this industry, or fragment it?

Predictions of mass consolidation have accompanied every software wave in forwarding — and the forwarder count never collapsed. What the deal market rewards today is measurable either way: buyers explicitly pay up for automation and AI-enabled operations, while undifferentiated forwarders stay valuation-constrained (AlixPartners Q1 2026 transport M&A review; PwC midyear deals outlook).

Here is why the answer matters less than it seems. If AI consolidates the industry, efficient books command the premium and inefficient ones get bought for their customer lists. If it fragments, efficiency is the ticket to staying independent at all. The two theses disagree about the industry and agree about you: the premium accrues to the adopter.

What’s real vs. what’s noise

The biggest AI story in freight this summer wasn’t a chatbot. Flexport built an AI agent for ocean market pricing. It ingests procurement rates, booking trends, blank sailings and market news, and forecasts where a lane’s price is heading in minutes instead of the two days the manual version took. In shadow testing, its forecasts came within 1% of the expert team’s. It now powers all of Flexport’s ocean pricing.

Notice what it doesn’t do: talk to customers.

That’s the pattern across every credible deployment we track, and across the forwarders we work with. The returns live behind the operator: document and email ingestion, reconciliation, the data entry that eats a morning before anyone speaks to a customer.

One vice president at a top digital-ingestion user told us she would “absolutely pay for” AI email classification. Another operation identified roughly $7,000 a month in AR/AP leakage that pattern-matching should have caught.

Customer-facing AI is a different story. The credibility backlash is already underway, with the most-cited statistic in this year’s AI discourse being MIT’s finding that 95% of GenAI pilots fail, and the first lawsuit over who owns AI-era freight data and workflows is in court. And the relationship — the phone call, the judgment call, the reason a shipper picked you over a portal — is precisely the part your customers don’t want automated.

The skeptic’s point deserves an answer, because it is real: sometimes the AI costs more than the person. True, which is why the test is never the demo. It’s the same arithmetic as everything else in this report. Did files per operator move? Did gross profit per file? Did conversion? If a tool can’t move one of those three numbers, it’s noise, whatever the demo looked like. Most firms holding back aren’t wrong about the tools they’ve seen, they’re just waiting for the version that shows up in the P&L. The giants just demonstrated what that looks like.

Three numbers to run

1. Gross profit per shipment, by mode: Total gross profit for a lane or mode, divided by shipments moved. Monthly. This is the giants’ yield problem, in your own book.

2. The yield check: Last year’s gross profit per container on a lane, minus this year’s, times this year’s volume. That’s what growth cost you. DHL’s ocean book ran this exact math in Q2: €32 less per TEU across 856,000 TEU while volume grew. Volume up, yield down is the most expensive way to feel busy. Run it on your top three lanes.

3. Conversion ratio, split by book: Operating profit divided by gross profit, never blended across business lines. DHL’s own retired disclosure is the cautionary tale: 23.8% for its forwarding unit, 17.5% once road freight was mixed in. A blended number is where a losing book hides.

Then audit yourself the way an acquirer would: check the quarter against the half-year, read your own footnotes, and back out anything that only happened once. 

Your scorecard

Benchmarks: Q2 2026 company filings, compiled August 2026.

The conversation continues

This report is a companion to our Logisyms panel, and there's more where this came from.