Generating Value with Data and AI

Artificial intelligence (AI) has become indispensable in many companies. Examples such as AI agents, early defect detection in manufacturing, or logistics process optimization show that AI is already being used across a wide range of business functions. Although AI is increasingly recognized as a key technology, sustainable value creation remains elusive in many organizations. Ambitious pilot projects are emerging everywhere, yet only a fraction make it into broad deployment – and thus into real value creation. The result is often sobering: companies pursue a multitude of activities without meaningful impact. AI applications quickly become a cost factor rather than a true value driver. More importantly, organizations miss out on significant potential because of ineffective AI use.

Many projects end as pilots

A recent survey of more than one hundred senior executives from various industries, conducted by Porsche Consulting in collaboration with the Fraunhofer Institute for Applied Information Technology (FIT), paints a clear picture: Success is limited not primarily by technology or algorithms, but by a lack of strategic direction, weak organizational anchoring, insufficient data quality, and corporate culture.

95 percent of respondents report active AI initiatives in their organizations. Yet only about 20 percent currently succeed in scaling AI applications beyond the pilot phase into widespread use. Projects launch and show initial promise but never reach steady-state operations. As a result, initiatives generate short-lived attention but fail to deliver sustained value for the organization.

Measuring the success of AI initiatives is equally critical. Overall, the results show that only 44 percent of companies quantify their economic contribution through key performance indicators (KPIs) such as AI return on investment. The gap becomes even more evident when comparing so‑called mainstream adopters (companies that introduce new technologies only once they are established) and innovators (companies that adopt AI very early). Only one quarter of mainstream adopters quantify the value contribution of their AI applications, while nearly 70 percent of innovators do so.

The issue is clear: if the impact of initiatives is not measured, they cannot be effectively monitored or steered. This leads to a pattern of high AI activity with limited impact. A combination that, due to lack of transparency, can result in misleading strategic conclusions.

Poor data quality, missing skills, and culture gaps

Many companies face structural barriers that significantly slow the scaling of AI applications. The survey highlights three main barriers that systematically limit long-term success: More than half of respondents cite poor data quality and lack of accessibility as the biggest obstacles. Large companies with revenues between 25 and 50 billion euros are particularly affected, with 86 percent reporting this challenge.

39 percent identify organizational silos and unclear responsibilities as key reasons why AI is not used enterprise‑wide. In many cases, departmental budgets determine the priority of use cases – not the business value. This results in local projects while cross‑functional applications receive insufficient support. Often, a central portfolio steering for AI applications is missing, leading to duplication of efforts, inefficient resource usage, and ultimately, project discontinuation.

Almost half of respondents list insufficient organizational skills as the greatest barrier. The survey underscores that early investments in AI competencies, a transparent culture, and empowering teams are essential to build trust in data. Without these foundations, AI initiatives typically fail to scale beyond pilots. Skills and cultural openness are decisive for the speed and scalability of AI efforts. 62 percent highlight qualified and empowered employees as the top success factor for effective AI use.

Success factors for value‑creating AI applications

The survey not only highlights key barriers. It also clearly identifies four levers companies can pull to overcome them.

1. Improve data quality and accessibility

Data availability and quality form the foundation of any successful AI initiative. Companies that recognize this treat data management not as an IT task, but as a strategic success factor. The challenge lies in consolidating and structuring data from diverse sources in a way that provides clear handling rules. Data governance becomes not a bureaucratic exercise, but a prerequisite for scalability. This approach ensures enterprise‑wide data consistency while enabling business functions to use the data essential for their AI applications efficiently.

2. Measure value consistently

The long‑term impact of AI depends above all on consistently measuring its value. However, these activities are often disconnected from strategic objectives. As a result, leadership support frequently falls short, preventing pilot projects from becoming sustainable initiatives. A fundamental shift in KPI logic is needed. KPIs should not focus exclusively on financial metrics but be embedded in a steering framework directly linked to strategic goals. In addition to classic performance and efficiency metrics, further measurement dimensions are essential. These include leadership effectiveness, employee satisfaction and engagement, real‑time customer enthusiasm, and innovation speed.

3. Ensure leadership and collaboration

The survey reveals that leadership support is one of the strongest levers for AI scaling. Both laggards (late adopters) and innovators agree unanimously that leadership backing and a clear data and AI strategy are critical success factors. However, implementation gaps remain. Only 52 percent of innovators report challenges in implementation. Leadership is central to AI success: it builds buy‑in across functions and embeds AI in daily work – not as an innovation project, but as the new standard. Actively engaged leaders also ensure collaboration and empower teams. One surveyed company highlights the high visibility of internal experts, particularly from IT, as a crucial factor in skill development. Practical training that showcases different tools helps lower barriers and increases acceptance in daily operations. Leaders who proactively build such structures significantly increase the chances of successful AI adoption.

4. Establish a product‑oriented approach

Leading organizations view AI not as a collection of isolated projects, but through the lens of stable structures with clear roles and responsibilities. 44 percent cite an established data and AI organization with defined roles and accountabilities as a key success factor – nearly twice as many as among mainstream adopters (23%). This shift from project logic to product and organizational logic is essential for ensuring sustainability and reuse. AI becomes not a temporary experiment, but an integral part of the business model. One participating company provides a practical example: it developed a model that maps the entire lifecycle of an AI use case – from idea to implementation – and integrates it seamlessly into end‑to‑end processes. Each application has a clear owner, a concise summary of its business value, and is tracked transparently and consistently through standardized tools. In this way, isolated initiatives evolve into a scalable AI portfolio with clear accountability, measurable impact, and high reusability.

Scaling AI instead of just testing it

Sustainable value creation with AI requires shifting from isolated experiments to a structured, enterprise‑wide approach. The formula for success is built on four core elements: high‑quality and accessible data, consistent value measurement, clear responsibilities, and a product‑oriented way of working.

The central question is no longer whether AI can create value, but whether the company is ready to manage AI with the same discipline it applies to its core processes. To do so, companies must understand how AI applications affect their value creation and how they reshape it in return. AI maturity is therefore not measured by the number of pilots, but by the number of value streams that are sustainably transformed.