Revlon’s ERP-driven manufacturing and fulfilment challenges, Metcash Australia’s Project Horizon cost escalation and execution risk disclosures together with one of our local retailer’s well-reported IT disruption.

Three continents. Three different organizations. One consistent signal. Large scale enterprise technology transformation rarely fails in the software itself. It fails in execution, when strategy, data, governance, operating discipline, and change readiness are not treated as co-equal workstreams alongside technology deployment.

This is not an argument against transformation. It is an argument for treating it as a more complex organizational change than many implementations allow for.

The numbers don’t lie

By some estimates, global IT spending is set to exceed 6 trillion dollars this year. Artificial intelligence, cloud platforms, and data centre infrastructure are driving much of this growth. In logistics and warehousing, where inventory, transport, procurement, and demand planning operate in tight interdependence, technology rollouts are increasingly system-wide events rather than isolated projects. In many cases, investment momentum is running ahead of organizational readiness.

Gartner estimates that more than 70 percent of new enterprise resource planning implementations may fall short of their intended business outcomes by 2027. Boston Consulting Group reports a similar pattern across digital transformation programmes globally. The issue is rarely technology capability. It is more often the gap between system ambition and organizational capacity to absorb change.

Bain’s recent research on generative artificial intelligence reinforces this point. Basic productivity gains are achievable. Sustained impact, however, tends to require changes to processes, controls, and operating models. Tools improve output. Systems determine outcomes.

The question is no longer whether Boards have artificial intelligence roadmaps. Most will. The more important question is whether leadership teams have the governance maturity and operational discipline to execute without eroding long term value in the process.

Where it actually breaks down

Across logistics and supply chain environments, familiar patterns continue to emerge. Many organisations approach major technology programmes as system upgrades rather than fundamental shifts in how the business operates. When operations, logistics, finance, and commercial teams are not deeply embedded in the design and execution, success is often measured by technical go live rather than end-to-end performance. As a result, delivery milestones are achieved, while outcomes such as inventory accuracy, fulfilment speed, and service reliability receive less sustained attention.

• Data quality is a persistent constraint across supply chains. These systems are only as reliable as the data they receive, yet migration rarely resolves underlying issues such as duplicate product codes, inconsistent supplier records, incorrect units of measure, incomplete customer information, or outdated inventory data. Once embedded in enterprise resource planning, warehouse management, and transport management systems, these problems become more visible, more interconnected, and more costly to correct.
• Implementation timelines also tend to underestimate operational complexity. Phased rollouts, pilot environments, data remediation cycles, and controlled cutovers are often treated as optional overhead, rather than essential safeguards that protect continuity and reduce risk during transition.
• Change management is frequently reduced to a communication exercise, rather than treated as a core part of operational transition. When new systems do not align with established workflows, teams tend to adapt informally to preserve continuity, but this can introduce parallel processes and gradually erode the intended benefits of the technology investment.
• Vendor selection is often shaped too heavily by demonstration environments, which rarely reflect the realities of live logistics operations. These “test” environments typically understate the complexity of exception handling, partial fulfilment, substitutions, reverse logistics, stock damage, delayed data feeds, and multi system integration challenges. As a result, misalignment often only becomes apparent after deployment, when operational pressure exposes gaps between design assumptions and real-world execution.

AI will not save you - it will expose you

Artificial intelligence is increasingly being layered onto logistics and enterprise systems. When implemented well, it can improve forecasting, strengthen exception handling, optimise replenishment, and support faster decision making.

But artificial intelligence does not replace the need for strong foundations. It raises the cost of their absence.

It will not correct underlying planning or data issues. In many cases, it will scale them more quickly across the organization. It can also make them less visible in the short term, as outputs appear more refined than the inputs that generate them. A system may appear intelligent while still being structurally fragile.

This is less a technology question than a leadership one.

What boards should be asking now

The central question is not whether an organization has an artificial intelligence roadmap. Most will.

It is whether leadership teams have the strategic clarity, governance discipline, operational capability, and organizational alignment required to deliver sustainable value across the supply chain.

In the era of artificial intelligence, data readiness is no longer a technical consideration. It is increasingly a board level risk factor and, in many cases, a source of competitive advantage.

The warehouse may be ready. The question is whether the organization around it is.

Ram Ramakrishnan, CEO of Cloud23

*this article does not reflect the opinions or views of Logistics News, Vicenda or their management and staff