I've been in enterprise technology for thirty years. I watched the shift from mainframe to distributed computing, from distributed to SaaS, and now from SaaS to AI. Each transition followed the same pattern: leaders recognized the wave was real, felt the pressure to act, and reached for the most visible lever available.
For most companies today, that lever is headcount reduction. Boards want ROI they can explain in a single sentence, CFOs want efficiency numbers that fit a slide, and cutting headcount is tangible in a way that's easy to defend. But if that's the primary lens through which your company is approaching AI, you are missing the larger opportunity by a wide margin.
Where the Real Money Lives
Here's a concrete example. We work with a client managing half a billion dollars in freight spend. If you eliminated half the people supporting that program, the savings would pale in comparison to what's available from analyzing service-level shifts, consolidation opportunities, and carrier optimization across their network. The difference is that the latter set of decisions can now be analyzed in seconds with the right data and platform, not over weeks with a team of analysts and an outside consultant.
Quality and Throughput Are the Right Measures
Go back to Kaizen. Look at the operational frameworks that have driven performance in manufacturing and logistics for decades. The two most important measures have always been quality and throughput. Improve both, and cost reduction will follow. That sequence matters. When headcount reduction becomes the goal itself, you've reversed the logic. You tend to cut functions that were carrying more weight than you realized.
I see this constantly. A company automates a workflow to eliminate the need for a person to do it. Six months later, they realize that person wasn't only executing the workflow. They were catching anomalies, maintaining carrier relationships, applying judgment in moments where the system alone would have gotten it wrong. That knowledge doesn't transfer to a software workflow. It's gone.
What we focus on at Trax, and what I tell every executive I work with, is a different question entirely. Not "what can we eliminate," but "what could this team accomplish if bandwidth weren't the constraint." Ten people doing the work of forty is a fundamentally different outcome than cutting five people to save a million dollars. The first creates compounding value. The second creates a one-time line item.
Start at the End
Here's something I'd put on a coffee mug: start at the end.
When you're evaluating where AI fits in your supply chain operations, don't begin with the technology or get pulled into implementation complexity before you've defined the goal. Start with what you're trying to accomplish.
In logistics, the goal is straightforward: move goods reliably, on time, at the lowest effective cost. Everything else serves that objective. So identify two or three use cases that feel out of reach given your current bandwidth and work backward from there. Which service levels across your shipping zones are costing you more than the delivery outcome justifies? Where do you have single-carrier concentration risk you haven't fully mapped? Which lanes have never been properly benchmarked against the broader carrier market?
These aren't theoretical questions. In our platform, analyses like these take seconds. The constraint was never the math. Rather, it was always the time and attention required to do it well.
The Instinct That Gets Leaders Into Trouble
There's a broader point here that I think about often, especially having watched several technology transitions up close.
I've watched people get left behind in every one of these shifts, and it's never because they couldn't learn something new. It's because they held on too tight to what made them valuable before. My father told me once to be careful what you ascribe value to, because whatever you decide is worth protecting, you will protect, even when that protection works against you. That instinct is powerful, and it runs through organizations as much as it runs through individuals.
When you ask your team what's possible with AI, understand that the answer you get is shaped by more than technical reality. People are working through real uncertainty about their own roles, and that colors what they tell you. As a leader, your responsibility is to understand enough about what's possible to separate the real constraints from those rooted in anxiety. The more fluent you become with your own data and its possibilities, the better equipped you are to have those conversations honestly.
The Question Worth Asking
Most CFOs want the same thing. Teams that surface insights faster, make better decisions with less friction, and find savings that weren't visible before. That is what a well-deployed AI strategy in logistics delivers.
The executives who come out ahead in this moment are the ones willing to ask harder questions, build genuine fluency with their data, and resist the gravitational pull toward the most obvious lever. The opportunity is real, and it's significant. It just requires starting in the right place.
If you're ready to find out what your freight data is telling you, connect with the Trax team. We work with global enterprises every day on exactly this problem
