Why AI Investments Aren't Delivering ROI-and How CEOs Can Fix It
Most companies implementing AI are automating the wrong things. They're layering generative AI into existing workflows designed for scarcity, then wondering why results disappoint.
The problem isn't the technology. It's the thinking behind how it gets deployed, according to Maine Pointe's leadership team, which recently published analysis on AI strategy in supply chain operations.
The ROI Question Misses the Point
Executives typically evaluate AI like any other technology investment: What's the return? That framework is backward.
"The fundamental issue is to have a value-centric mindset," says Avik Ghosh, head of AI and product innovation at Maine Pointe. "You have to ask yourself: what is the fundamental value of this process? Why does this process exist?"
Companies seeing strong returns aren't asking "how do we use AI?" They're asking "what are we actually trying to achieve?" and rebuilding from there.
Joseph Esteves, CEO of Maine Pointe, describes the common mistake: "We put AI into an inefficient system, and we got inefficiency out of it. We were just automating constraints rather than removing them."
What Real Impact Looks Like
Once constraints are removed instead of automated around, business impact accelerates. Decision cycles compress. Risk visibility improves. Teams scale improvements rapidly.
But speed depends on one factor: "Gen AI's ROI is not capped by technology's potential," Esteves says. "It's capped by how quickly organizations can evolve the way their people work."
The investment extends beyond software. CEOs must redesign workflows, retrain existing staff, and create roles that don't yet exist.
Inverting Organizational Structure
High-performing companies are restructuring around AI differently. Ghosh describes a model gaining traction: flipping the traditional hierarchy.
"Instead of the CEO at the top, that position is now at the bottom," he says. "The philosophy moves from who reports to you, to who do you support. The root supplies nutrients to the leaves-the employees-who can grow in their own ways."
Esteves frames the stakes plainly: "Democratize innovation. Get it in everyone's hands. Give everyone the chance to solve problems, because when you scale that way, you can have hundreds, thousands solving problems the top of the organization didn't even know existed."
Timing Matters Now
For supply chain leaders managing tariffs and geopolitical shifts, agility has become survival. Traditional sourcing events took a year to complete; companies then locked in contracts for two to three years.
"Now, with all this uncertainty, you don't have that luxury," Ghosh says. "AI enables you to be more agile to deal with changing macroeconomic and geopolitical realities in real time."
For executives evaluating AI strategy and consulting partners, the question has shifted. Don't ask what AI can do. Ask what your organization is actually trying to achieve-then rebuild around that answer.
Learn more about AI for Executives & Strategy, or explore the AI Learning Path for CEOs.
Your membership also unlocks:
