AI and ROI: 5 strategic priorities for EMEA companies

Organizations in Europe, the Middle East, and Africa find themselves today at a critical juncture in their artificial intelligence adoption journey. A new IBM study reveals that 66% of companies in the EMEA region have already achieved significant productivity improvements thanks to AI, yet only 20% have actually reached their return on investment (ROI) objectives. This discrepancy highlights a fundamental challenge: how to transform AI’s tactical benefits into strategic and lasting value? The answer lies in five strategic priorities that organizations must implement to accelerate their path toward concrete and measurable ROI.

The current context: opportunities and challenges

IBM’s “The Race for ROI” study, conducted in collaboration with Censuswide in September 2025, involved over 3,500 business leaders, including 200 Italian senior leaders. The data that emerged paints a picture of a market in transformation, where AI is no longer a simple technological promise, but an operational element integrated into the strategies of the most advanced organizations. In Italy specifically, 56% of the organizations interviewed declared significant improvements in productivity thanks to AI, while 13% report that AI benefits have influenced the transformation of their business model. This data is encouraging, but also represents an opportunity that has not yet been fully exploited. The most evident impacts of AI are concentrated in specific sectors: logistics records the highest rate of operational efficiency increase (75%), followed by automotive and transport (67%), manufacturing (60%) and energy and utilities (60%). Among the most cited impacts of AI are time savings (45%), cost reduction (41%), employee satisfaction (42%), revenue increase (37%) and Net Promoter Score improvement (43%). These benefits demonstrate that AI is not a one-dimensional tool, but creates value through multiple operational and strategic channels. However, as emerges from IBM’s report, significant inequalities persist in the ability to realize value from AI. 72% of large enterprises declared an increase in productivity, while only 55% of SMEs and 55% of public organizations report the same result. This gap of 17 percentage points between large companies and SMEs reveals a critical situation: small and medium enterprises, which constitute the backbone of the European and Italian economy, remain behind in the AI ROI race.

The future vision: AI agents

Particularly significant data concerns AI’s future prospects: 92% of leaders expect measurable ROI within two years of adopting agentic AI. Artificial intelligence agents – autonomous systems capable of making decisions and taking actions based on defined objectives – represent the next level of evolution compared to current chatbots and generative AI tools. This optimistic prediction (92% versus the 20% that has already achieved ROI today) suggests that leaders see agents as a significant qualitative leap in terms of impact and value generation. Agents could, for example, autonomously manage complex processes like supplier negotiation, supply chain optimization, or advanced data management, with minimal human involvement.

The five priorities

The IBM report outlines five priorities to accelerate Artificial Intelligence ROI:

establish an effective operational model for AI;
cultivate AI literacy and a culture of innovation, from the board of directors to all levels;
become comfortable with uncertainty and rapid changes;
understand the risks related to AI implementation;
establish a cross-company “AI Board” to mitigate risk.

1. Integrated AI vision

The first critical element on which to build AI success is the establishment of a solid and universally understood operational model for the entire organization. This doesn’t simply mean having an AI solution; it means creating a common approach to transformation that involves all business functions with a clear definition of ownership and responsibilities. A transition from isolated initiatives to a coherent and shared strategy is necessary: too many organizations see AI as a technological appendage managed by a single team, rather than as a transformation that must permeate the entire business structure. Creating this integrated vision first requires shared governance: IT, data, compliance, legal, operations and various business units must sit at the same table from the beginning. These are not bureaucratic committees, but operational coalitions that make rapid decisions on who authorizes projects, how they are funded, and how each department’s objectives align with common AI goals. Research highlights that about 70% of AI implementation failures concern people and organizational processes, not the technology itself. This data is revealing: it means that having the world’s most sophisticated algorithm is useless if the organization doesn’t know how to welcome it, manage it, and make it evolve. Companies that manage to create this shared vision succeed in translating tactical benefits into lasting strategic value.

2. Widespread literacy and innovation culture

If AI will be ubiquitous in work in the coming years, it cannot remain the prerogative of a niche of experts. This second strategic element concerns the democratization of knowledge: from the C-suite to operational employees, everyone must develop an authentic understanding of how AI works and, more importantly, how it can transform their daily work. This doesn’t mean turning everyone into data scientists. It means creating structured training programs at different levels: strategic sessions for leaders who must make decisions about AI investments; practical training for teams that will use the tools daily; general awareness sessions for the rest of the organization. The goal is to create a culture where curiosity toward AI is encouraged, not feared. Also crucial is actively managing narratives and perceptions. When communicating an AI project, the message matters more than the technology itself. Communicating that AI frees up time for more meaningful work, not that it replaces employees, radically changes the adoption rate. Organizations that manage this communication empathetically and transparently experience a significantly higher adoption rate.

3. Ability to navigate uncertainty and rapid change

In the next decade, uncertainty will not be a temporary exception but the structural norm. AI will evolve rapidly, new tools will emerge continuously, regulations will change, and competitors will try increasingly different approaches. This third strategic element requires rethinking organizational culture profoundly. Organizations must develop “a culture that embraces change and uncertainty and fosters rapid and targeted innovation.” This is a profound cultural shift from traditional business models that seek to minimize uncertainty through rigid planning.

4. Conscious and proactive risk management

The fourth strategic element transforms risk management from a bureaucratic cost to a competitive advantage. Too often, organizations see risk management as a brake on innovation. Actually, when done well, it’s exactly the opposite. Implementing AI without a rigorous understanding of associated risks – regulatory, reputational, operational, ethical – is like driving at night without headlights. It means that sooner or later you’ll have an accident, and when it happens, it will be costly and public. Organizations that instead understand and proactively manage these risks build internal and external trust, attract better talent, and better meet emerging regulatory requirements. This means establishing rigorous audit processes on AI systems, traceability of automated decisions, active involvement of compliance and ethics functions in project evaluations, and continuous monitoring of model performance and behavior. It’s not something you do once and then forget; it’s a permanent vigilance activity.

5. Create a hybrid governance body

The fifth element emerges from a simple but important recognition: decisions about which AI projects to pursue cannot be made only by technicians, and not even only by executives. They require a multidisciplinary perspective that simultaneously considers technical feasibility, business impact, ethical implications, regulatory compliance, and organizational sustainability. For this reason, organizations should establish a governance body – as IBM calls it an “AI Board” – that brings together representatives from business, technology, compliance, ethics, and leadership. The role of this structure is not slow bureaucracy; on the contrary, it should facilitate fast innovation within clear boundaries. How should it work in practice? Business units identify AI use cases and bring them to the Board for rapid evaluation. The Board examines: is it technically feasible? Does it create business value? Does it respect our ethical principles? Is it compliant with regulations? If it passes all these checks, it receives the green light and proceeds quickly. This approach allows organizations to move quickly without losing control. Combined with the second pillar (raising AI awareness throughout the organization), this hybrid governance model gives individual business units and operational teams a high level of autonomy to act safely, knowing there is a solid ethical and operational framework supporting them.

Critical barriers to AI scalability

The two main barriers to scalability are legitimate concerns about security, privacy and ethics (66% of leaders), and the technical complexity of integration with legacy systems (65% of leaders). Many European organizations have infrastructures built over decades: integrating AI requires specialized skills, complex projects, and risks of operational disruption. These obstacles are not insurmountable, but require a methodical approach that simultaneously addresses risks and technical complexity, with clear governance and transparent communication.

Opportunity in uncertainty

The path to artificial intelligence ROI is neither linear nor fast. However, organizations that implement the five strategic priorities outlined here – organizational integration, widespread literacy, ability to navigate uncertainty, conscious risk management, and hybrid governance – will be in a privileged position to capitalize on the technology’s value. In the Italian context, where 56% of organizations have already seen AI benefits but only 13% have transformed their business model, there is still a window of opportunity to become the first innovators of these methodical approaches. Although SMEs are at an earlier stage in their adoption journey, they represent the segment with the greatest growth potential: by adopting these strategic priorities they can significantly accelerate their path toward concrete and sustainable ROI.

Sources: