Supply Chain Technology 2025–2026: What Worked, What Flopped, and What Leaders Are Doing Differently

Not every supply chain technology initiative that launched in 2025 delivered results, and the leaders gaining traction in 2026 are doing things differently.

On May 27, supply chain executives from Accenture, e2open, and Supply Chain Movement joined our Thought Leadership Webinar Series for a candid conversation: Supply Chain 2025–2026: Tops or Flops and the Lessons Shaping What Comes Next. The format was simple: six statements, each one evaluated as a "top" or a "flop", designed to cut through the noise and get to what's actually happening.

The panelists brought decades of hands-on experience between them: 

Marta Diaz Lorenz, Managing Director at Accenture Spain, with 20 + years in supply chain transformation across industries

Pablo Quintano, AVP of Industry Solutions at e2open, a WiseTech Global company

Martijn Lofvers, CEO and Chief Trendwatcher at Supply Chain Movement, host of the session

Here's what they said, and what it means for 2026 supply chain planning.

Statement 1: "AI-first supply chain projects were truly reshaping operations in 2025"

Verdict: Flop

Since analytics, machine learning, and rule-based optimization all got branded as AI, the “AI-first” label lost meaning. It made it impossible for organizations to set expectations or measure progress of AI-first supply chain projects.

According to an Accenture analysis, roughly 95% of organizations that recently implemented AI didn’t achieve expected outcomes, and agentic AI adoption in supply chains sits under 10%, according to a Boston Consulting Group (BCG) report from 2025. Pablo put it plainly: "We still see many companies three steps below where they need to be to adopt these new technologies.”

The second issue was the gap between ambition and execution. Much of 2025 was made up of companies asking, "what can we actually do with AI?" Companies tested reliability, piloted narrow use cases, and worked with limited user groups. Broad operational reshaping wasn’t there yet.

Both panelists agreed that the root issue with “AI-first” is the operating model. AI adoption requires changes in mindset, roles, and ways of working that most organizations haven't fully addressed. And then there’s a practical constraint that often gets overlooked: AI needs data infrastructure to be effective. Companies need a data layer that extracts and consolidates inputs from multiple systems. While you don't need a complete data environment to start, you do need to identify areas and data sets where you can build traction.

The takeaway: Don't ask "how do we go AI-first?" Ask "what business outcome are we trying to achieve, and how can AI play a role in getting there?"

Statement 2: "In 2026, AI-first organizations are finally embedding AI at the core of supply chain decision-making"

Verdict: Split (Marta — Top; Pablo — Flop)

“AI-first organizations are finally embedding AI at the core of supply chain decision-making" generated an interesting disagreement between Marta and Pablo.

Marta called it a top, with a nuance. AI has climbed to the forefront of nearly every supply chain conversation. Companies are building roadmaps that explicitly incorporate AI, and the question around AI has moved from "should we use it?" to "where do we start?"

Pablo disagreed and called it a flop. In his experience, many isolated AI pilots deliver little return, and the companies getting value are the ones treating AI as part of a broader transformation vs standalone technology. "Remove the hype around AI-first, AI-only transformations," he said. "Put AI in the context of a broader discussion."

He also raised a point: supply chain connectivity must come before AI. If a company struggles to get suppliers to confirm purchase orders or takes six months to onboard a new trading partner, applying AI won’t fix the root cause. Solve the connectivity problem first, and you’ll get better data, which makes AI actually useful.

The takeaway: AI adoption is accelerating, but the organizations making it stick have already built the supply chain connectivity and data infrastructure to support it.

Statement 3: "The vendor/system integrator/client triangle drove real transformation wins in 2025"

Verdict: Mixed 

For companies who successfully implement a supply chain technology transformation, responsibilities are clearly defined, there’s a shared vision, and all three parties — the technology vendor, the system integrator, and the client — play active roles throughout the engagement.

Like a three-legged stool, each of these three parties brings distinct value and support to the supply chain transformation process. The vendor brings deep product expertise and capability. The system integrator brings market knowledge, best practices, and change management, and the client brings business knowledge and decision authority.

This three-way partnership model is required for complex, AI-enabled supply chain transformations. As Pablo put it: "This is the only way things are going to happen." The question is whether the operating model and governance structures are ready to absorb it.

Meanwhile, failure happens when technology implementations accumulate customizations because no one said no. These "Frankenstein implementations" are expensive to maintain and difficult to evolve. The fix? It comes from better discipline around what's standard, what's configurable, and what doesn't need to be built at all.

The takeaway: Transformation at scale requires all three parties to show up as active contributors with clear ownership.

Statement 4: "Organizations had a supply chain roadmap and followed it in 2025"

Verdict: Flop

Even sophisticated organizations still struggle with siloed thinking. Logistics teams avoid planning conversations. Global trade teams run their own roadmap, separate from the rest of the supply chain. The result is parallel workstreams that never connect or compound.

The macro environment made this misalignment worse. Tariff volatility, pre-loading activity ahead of US policy changes, and geopolitical uncertainty caused many companies to pause, and lose momentum on initiatives in flight. And new capabilities emerged so quickly that roadmaps built six months ago were already outdated.

However, the forward-looking outlook is more optimistic. The failed pilots of 2025 are driving teams to reassess priorities and realign execution before charging forward. More cross-functional conversations are happening between supply chain, finance, and commercial teams to align around a shared roadmap, and S&OP and IBP processes are increasingly converging with FP&A cycles.

"The leaders making faster and more solid transformations are the ones with a clear, cross-functional roadmap," Pablo said. "It's easier to infuse new initiatives into a roadmap that already exists and is understood across the organization."

The takeaway: A supply chain roadmap that lives in one function is not a supply chain roadmap. Cross-functional alignment, especially with finance and commercial, is what separates vision from execution.

Statement 5: "Digital twins, control towers, and predictive platforms broke through in 2025"

Verdict: Mixed

There was real investment in digital twins, control towers, and predictive platforms in 2025, but the results were uneven, and adoption fell short of expectations.

Digital twins

For digital twins, ambition outran execution. One forecast predicted that 40% of companies would integrate digital twins into operations by 2025. But hiring for digital twin skill sets stagnated at roughly 22% of supply chain organizations, about half what was expected.

Control towers

The term "control tower” still means different things. For some, it's a KPI dashboard. For others, it's an active exception management mechanism. The version of a control tower that delivers strategic value provides root cause analysis, impact modeling across supply chain tiers, and actionable recommendations. But for some organizations, this is still more aspiration than reality.

Pablo has worked on the concept of connected, cross-functional control towers for the last 15 years. The vision for what a control tower should do remained the same: to see, understand, and act. But the key challenge has been the ability to connect the data across the end-to-end supply chain.

Predictive platforms

You need clean data, partner connectivity, and a cross-functional operating model for predictive platforms to run their engine without stalling out.

The key enabler for emerging supply chain technologies to stop being proof-of-concept projects and start delivering enterprise-wide impact, is connecting supply chain visibility to financial outcomes. When a digital twin or control tower can show commercial and finance teams what a supply chain decision means for service levels, cost, and revenue, it earns a seat at the enterprise table.

The takeaway: The technology for cross-functional, predictive supply chain visibility is mature enough to deploy. The gap is still in data connectivity and the organizational willingness to act on what the tools surface.

Statement 6: "Autonomous supply chains moved past exception-based management in 2025"

Verdict: Flop

The panelists agreed: autonomous is a premature label. Humans are still largely running the show.

In many organizations, even moving to exception-based management, where planners accept system recommendations and only intervene on flagged exceptions, is still a significant cultural shift. Many supply chain teams still manually adjust nearly every output from their planning tools, regardless of what the system recommends. Getting organizations to trust the algorithm is meaningful progress, but we’re nowhere near full autonomous operations.

Rather than chasing fully autonomous supply chains, organizations should view AI agents as support tools that handle data gathering, pattern recognition, and routine recommendations. That in turn frees human planners to focus on interpretation, judgment, and decisions that require context the algorithm doesn't have.

Two human capabilities that were flagged as increasingly valuable in this environment: cross-functional intuition and translating data into insight. Experience and interpretation make human expertise difficult to replace.

The takeaway: Stop optimizing for "autonomous." Focus on the human-AI partnerships that help teams make faster, better decisions.

The big picture: Three things that will separate leaders from laggards in 2026

Across all six statements, a few consistent themes emerged.

The "power of three" is not optional. Technology vendors, system integrators, and clients each have a role in supply chain technology transformation, with clear ownership, and a shared vision.

AI needs context, not just a budget. AI works when it's embedded in a transformation roadmap tied to specific business outcomes. The organizations that saw return from AI in 2025 were the ones that asked, "what are we trying to achieve?" before asking "what can AI do?"

Data and connectivity are the foundation for AI. You don't need perfect data to start, but you do need to know where gaps exist, and to extend your data connections beyond your four walls to your upstream suppliers, and downstream logistics partners, and channels. External connectivity is what turns AI from a reactive tool into a proactive one. Without it, even the most sophisticated tools just help you to react to problems you could have seen coming.