Investment in artificial intelligence is skyrocketing, but most companies are not seeing the benefits they hoped for. A recent global survey by the Boston Consulting Group (BCG) shows that only 5 percent of the companies surveyed are currently achieving AI value on a large scale, even though almost all of them are already investing in it. At the same time, 60 percent report hardly any measurable added value, despite considerable expenditure.
In its Build for the Future studies for 2024 and 2025, BCG surveyed more than 1,000 and around 1,250 CxOs and senior executives from a total of 68 countries and several core industries. The industries surveyed include consumer goods, financial services, manufacturing, technology, and healthcare. The studies are based on a standardized AI maturity model, which was expanded from 30 (2024) company-wide capabilities to 41 (2025) capabilities. The 2025 analysis shows that 35 percent of companies are at the “scaling” stage and 5 percent are in the top category, “future-built.” Together, this means that 40 percent are already rolling out AI on a large scale. At the same time, 60 percent of companies are achieving little or no material AI value and are considered laggards.
The vicious circle of laggards
Leading companies set a spiral in motion. They generate measurable AI value and reinvest the resulting profits in further capabilities, talent, and technologies. According to BCG, Future-Built companies achieve 1.7 times higher revenue growth, 3.6 times higher three-year total shareholder return (TSR), 2.7 times higher return on invested capital (ROIC), and 1.6 times higher EBIT margins than laggards. They also file 3.5 times more patents.
These results feed into a self-reinforcing dynamic. Future-built companies plan to increase their IT spending by 26 percent more than laggards and allocate up to 64 percent more of their IT budget to AI. In areas where they use AI, they expect revenue growth to be about twice as high and cost reductions to be 1.4 times greater than laggards. Companies without measurable AI value thus risk falling further and further behind, both technologically and economically.
70 percent of the value lies in the core business
A common misconception: Many companies focus primarily on AI applications in support functions such as HR, IT, or legal. However, BCG analyses show that most of the potential lies in core business. According to BCG, around 70 percent of AI value in 2025 will come from core functions such as research and development, sales, marketing, production, supply chain, and pricing. R&D and innovation alone account for 15 percent of the total potential. This is followed by digital marketing (9 percent), customer journeys (8 percent), sales (7 percent), and production (9 percent). Support functions contribute around 30 percent to AI value. This represents a further shift from 2024, when the share of core business was still 62 percent.
The role of IT is striking: its share of perceived AI value has risen from 7 percent (2024) to 13 percent (2025). This represents an increase of 6 percentage points. Successful companies are not only investing “more in IT,” but are also building a scalable, modular, and interoperable technology architecture. This architecture includes central AI platforms, reusable building blocks, and a uniform data foundation. Future-built companies are significantly more likely to rely on central AI platforms, standardized data models, and group-wide data policies than laggards.
Concrete effects are already visible: A North American telecommunications provider used AI to analyze call center conversations. This reduced the average interaction time by 20 percent, reduced transfers to live agents by 25 percent, and shifted around 30 percent of contacts to AI-supported chatbots. This enabled it to achieve a cost reduction of around 25 percent in the unit concerned.
Agent AI to double its value contribution by 2028
A new accelerator is Agentic AI, i.e., autonomous AI agents that can observe, reason, plan, use tools, and act. By 2025, agents already accounted for around 17 percent of total AI value. By 2028, this share is expected to rise to around 29 percent. At the same time, BCG expects the value share across the three forms of AI to shift from around 45 percent generative AI, 38 percent predictive AI, and 17 percent agentic AI in 2025 to 38 percent generative, 35 percent predictive, and 29 percent agentic AI in 2028.
In this logic, these agents are not simply additions to existing processes. In many cases, they require a fundamental redesign of workflows in which human and digital “employees” work together. Future-built companies already invest around 15 percent of their AI budget in agents and use them significantly more often than others.
According to BCG, a total of around 46 percent of companies are experimenting with or piloting agents. Around 16 percent of them are already reporting measurable added value. A leading manufacturer of electronic devices orchestrates AI workflows in more than 200 factories via a central platform for agent-based solutions – a so-called “company store.” The modeled EBIT potential: over $300 million.
People, Processes, Technology
The BCG studies come to a clear conclusion: around 70 percent of the hurdles in AI transformations relate to people and processes. Around 20 percent relate to technology and only around 10 percent to algorithms. Often, there is a lack of clear priorities, robust business cases, acceptance within the organization, or the ability to truly rethink existing workflows. On the technology side, issues such as data quality, integration into existing systems, security, and responsible use of AI dominate.
BCG therefore recommends the 10-20-70 rule: focus around 10 percent of resources on algorithms, 20 percent on technology and data, but 70 percent on people and processes. The research also shows that AI leaders, i.e., companies in the top two maturity levels, invest significantly more in digital and AI capabilities. They also allocate more FTEs to digital and AI and train a larger proportion of their workforce in AI skills than less mature companies. The 2025 data also shows that future-built companies want to empower more than half of their employees in AI. Laggards plan to do so for only about one-fifth on average.
A practical example is provided by a large international bank that redesigned its HR function with an AI-first approach from “hire to retire.” It defined key employee journeys such as “I apply and join,” “I get feedback,” and “I get rewarded,” and used these to derive a skills roadmap. It linked this to clear KPIs such as time-to-hire, 90-day retention, and service request turnaround times.
What sets Future-Built companies apart and how to catch up
BCG defines “future-built” as companies that are at the forefront of AI innovation. These companies have built advanced AI capabilities across functions and consistently generate substantial business value. In the 2024 study, fintech (around 49 percent), software (around 46 percent), and banking (around 35 percent) have particularly high proportions of AI leaders (a combination of “scaling” and “future-built”). More than half of these top performers are established incumbents, not purely digital natives. Examples range from an automotive manufacturer that uses GenAI to create tender documents around 50 percent faster and accelerate bid analysis, to a global financial institution that is aiming for productivity gains of around $1 billion by 2030.
The Five-Point Playbook of the Pioneers
A consistent five-point playbook can be derived from the two studies, which Future Built companies are already following:
- Multi-year, top-down anchored AI ambition
AI is managed as a strategic program with a clear value agenda, roadmap, and dedicated funding. Nearly all Future-Built companies report highly committed C-suites, compared to a few percent among laggards. - Focus on high-value workflows instead of arbitrary use cases
Leading companies prioritize a few highly effective workflows and rely on a clearly tracked roadmap. In future-built companies, around 62 percent of prioritized initiatives have already been deployed, compared to only around 12 percent among laggards. - Building solid IT and data foundations
More than 50 percent of future-built companies operate on a unified, company-wide data model, compared to only around 4 percent of stagnating companies. They also operate centralized AI platforms and enforce data policies significantly more often. - Parallel development of critical skills with a focus on people and processes
Future-built companies are several times more likely to engage in strategic workforce planning for AI, provide structured learning time, and actively involve employees in the co-design of AI solutions. - Rigorous tracking and governance throughout end-to-end transformation
More than 60 percent of Future-Built companies systematically track AI value contributions, compared to only 17 percent of stagnating companies, and anchor responsible AI with clear guardrails.
Those who consistently implement these principles and take the 10-20-70 prioritization to heart can gradually expand their AI maturity and leverage real business value on a large scale. A good starting point is to take an honest look at the status quo, for example, through an AI maturity assessment. This will result in a clearly prioritized roadmap, from initial, easily financeable flagship workflows to the redesign of entire functions with an AI-first operating model.
About the 2024 study: The BCG study “Where’s the Value in AI?” (October 2024) is based on a survey of over 1,000 CxOs and senior executives worldwide. Participants assessed their AI maturity across 30 capability dimensions. The survey covered more than 20 industries in 59 countries and was conducted in the third quarter of 2024.
About the 2025 study: The BCG follow-up study, “The Widening AI Value Gap: Build for the Future 2025” (September 2025), expanded the data set to include 1,250 CxOs and senior executives with AI decision-making authority from 68 countries in Asia, Europe, and North America. The maturity model was expanded to 41 capability dimensions. The survey covers nine core industries and was conducted in the second quarter of 2025.
Cover image: © Focus
Quellen:
BCG-Studie 2024: „Where’s the Value in AI?“: https://www.bcg.com/publications/2024/wheres-value-in-ai [Februar 2026]
BCG-Studie 2025: „The Widening AI Value Gap: Build for the Future 2025”: https://www.bcg.com/publications/2025/are-you-generating-value-from-ai-the-widening-gap [Februar 2026]