Enterprise organizations are investing billions in artificial intelligence and enterprise resource planning (ERP) systems, yet most initiatives fall short because they rely on outdated organizational structures that can't support modern technology. According to new research from consulting firms and enterprise service providers, the problem isn't the technology itself; it's that companies are trying to run sophisticated digital platforms through traditional, siloed operating models that slow decision-making and fragment execution.

Why Are So Many AI and ERP Projects Failing?

The numbers tell a sobering story. Fewer than half of ERP programmes achieve or exceed their intended business outcomes, and only around 30% of ERP transformations deliver sustainable change while meeting their target objectives. The situation with AI is similarly challenging. While 62% of organizations are now applying AI to core business and operational processes, enterprise readiness continues to lag adoption. Only 40% of organizations have an enterprise AI strategy, and just 20% extend it across their broader ecosystem.

The gap between ambition and execution stems from a fundamental mismatch. Organizations invest in cutting-edge systems while maintaining traditional functional hierarchies, unclear decision rights, and siloed teams. This creates what consultants call a "one-off programme" rather than a lasting organizational capability. When technology implementation happens without corresponding changes to how work flows through an organization, transformation stalls.

What Organizational Changes Do Companies Need to Make?

Successful enterprises are fundamentally rethinking how work moves from concept to delivery. Rather than organizing around functional departments, leading organizations are building cross-functional teams centered on specific business outcomes. This structural shift reduces unnecessary handovers, strengthens accountability, and enables faster execution.

Decision-making speed has become a critical competitive advantage. Many organizations across the Middle East and globally benefit from strong executive sponsorship and robust governance structures, but unclear decision rights cause issues to circulate through multiple committees before action is taken. Faster decision-making helps organizations maintain momentum, reduce rework, and accelerate value creation.

How to Build a Sustainable AI and ERP Operating Model

• Establish Clear Ownership: Define explicit ownership of business outcomes rather than spreading accountability across multiple departments or committees. This ensures someone is responsible for results and can make decisions quickly without endless escalation.
• Design Cross-Functional Teams: Organize delivery around specific business outcomes using teams that span traditional functional boundaries. This reduces handovers, strengthens accountability, and enables faster, more effective execution of transformation initiatives.
• Create Explicit Decision Pathways: Define clear decision-making processes and authorities so that issues don't circulate through multiple committees. Faster decision-making has become a significant competitive advantage in scaling AI and ERP initiatives.
• Embed Governance Into Portfolio Operations: Rather than treating each transformation as a standalone initiative, institutionalize consistent governance, decision-making, and delivery practices across the entire organization to make digital transformation a repeatable capability.

Beyond individual projects, organizations should embed these principles into a broader portfolio operating model. This transforms digital transformation from a series of isolated programmes into a repeatable organizational capability that continues to improve long after go-live.

What's Constraining Enterprise Execution Capacity?

The real bottleneck isn't funding; it's talent and leadership bandwidth. Many organizations attempt to deliver too many transformation initiatives simultaneously, stretching internal expertise and leadership capacity too thin. According to research from CGI based on discussions with more than 1,800 business and technology executives, nearly 70% report difficulty recruiting IT talent, and 52% indicate that talent shortages materially impact programs and execution capacity.

Cost pressure remains the number one constraint facing organizations, while 45% of executives say legacy systems significantly challenge their data and AI strategies. When organizations fail to align project ambitions with realistic delivery capacity, progress becomes fragmented, with numerous initiatives launched but relatively few completed successfully.

"Technology can unlock transformation, but only the right operating model can sustain it," said Stefan Westdijk, Managing Partner at Argon & Co.

Stefan Westdijk, Managing Partner at Argon & Co.

To address these constraints, C-level executives are increasingly shifting toward substantial and selective managed services models to strengthen delivery capacity and support scalable AI-enabled transformation. Clients are consolidating toward fewer, trusted partners capable of combining business consulting, systems integration, and digital reengineering to deliver end-to-end outcomes.

How Can Organizations Retain Strategic Control During Transformation?

While specialist implementation partners play an essential role in large-scale transformation programmes, organizations risk losing strategic control if vendors begin driving product roadmaps and key business decisions. Leading organizations are retaining ownership of products, priorities, and outcomes internally. This approach preserves accountability, strengthens internal capabilities, and ensures critical organizational knowledge remains within the business long after implementation is completed.

The shift toward agentic AI, which uses autonomous software agents to automate routine tasks and coordinate workflows, is creating new opportunities for organizations that have modernized their foundations. Kyndryl and Amazon Web Services (AWS) recently expanded their strategic collaboration to help customers adopt and scale agentic AI across mission-critical workloads. According to the Kyndryl Readiness Report, more than 68% of customers are investing heavily in AI, but most aren't realizing the anticipated benefits or operational efficiencies.

"Many organizations are focused on adopting agentic AI, but they are stuck in the experimentation phase instead of applying it in a way that actually makes a difference for their business," said Giovanni Carraro, Global Strategic Alliances Leader at Kyndryl.

Giovanni Carraro, Global Strategic Alliances Leader at Kyndryl

The path forward requires organizations to view operating model redesign not as a side project but as the central lever for transformation. When delivery is built into the organization rather than managed around it, AI and ERP initiatives become sustainable. This means establishing clear ownership of business outcomes, defining explicit decision-making pathways, organizing delivery through cross-functional teams, and embedding these principles into a broader portfolio operating model that continues to improve over time.