Business

Beyond traditional metrics: Rethinking the ROI of AI in 2025

"Companies that rely only on traditional ROI no longer even see the tip of the iceberg of the value of AI." McKinsey documents the winning approach: 70% investment with predictable ROI, 20% strategic innovation, 10% breakthrough exploration. Benefits manifest in cycles-optimization (0-12 months), reinvention (1-2 years), disruption (2+ years). 83% of Fortune 500 use digital twins to simulate impact. The debate is no longer metrics vs. strategy: it's who has integrated frameworks vs. who loses relevance.

While our previous article focused on ROI measurement for artificial intelligence-based SaaS technologies, this updated contribution presents an evolved perspective: companies must complement the accuracy of traditional ROI calculations with a multilevel strategic vision. This approach is no longer an option, but a competitive imperative in the digital ecosystem of 2025.

The challenge of evaluation: balancing immediate results and long-term transformation

The reality is well established: evaluating AI investments solely through traditional ROI metrics is inadequate and short-sighted. Organizations that limit themselves to this approach are already losing ground to competitors with a more strategic vision.

"Companies that fail to look beyond immediate ROI are not simply missing transformative opportunities, they are actively undermining their future relevance," says Accenture Chief AI Officer Sarah Chen, recently interviewed at the World Economic Forum 2025 [1]. "It's not about abandoning ROI, it's about integrating it into a more sophisticated evaluation framework."

The most recent studies in behavioral economics by the Harvard Business Review (March 2025) confirmed that organizations still tend to favor immediate benefits over investments with potentially exponential but more distributed returns over time [2]. This cognitive trap has become particularly dangerous in the era of generative AI, where the most significant competitive advantages often emerge only after initial phases of apparent limited returns.

Integrating ROI with strategic perspectives: the new 2025 standard

1. Balancing optimization and disruptive innovation

AI adoption driven solely by ROI inevitably leads only to incremental improvements. The McKinsey Global Institute's "AI Investment Strategies 2025" report shows that leading companies have adopted a "70-20-10" approach: 70 percent of AI investments for optimizations with predictable ROI, 20 percent for medium-term strategic innovations, and 10 percent for potentially revolutionary explorations [3]. This balance has become essential to maintain competitiveness in increasingly volatile markets.

2. Enhancing augmented collaborative intelligence

Traditional systems continue to perpetuate information silos that stifle innovation. According to a February 2025 MIT Technology Review study, today's AI platforms not only break down these barriers, but actively create new models of human-machine collaboration that generate exponential value [4]. Most advanced investment evaluations now include specific indicators of "collaborative intelligence" that measure this transformative potential.

3. Building systemic adaptability, not just efficiency

In an environment of increasing unpredictability, the Deloitte AI Resilience Report 2025 highlights how leading organizations value AI not only for its efficiency under normal conditions, but for its ability to adapt quickly to disruptive scenarios [5]. AI-based stress analysis has become a standard for assessing organizational resilience. Companies that ignore this dimension in their assessments are drastically underestimating the strategic value of AI.

4. Orchestrating the expanded digital ecosystem

The economies of 2025 function as hyper-connected ecosystems. Forrester's research "AI-Driven Business Ecosystems" (April 2025) shows that AI solutions not only generate value within the organization, but redefine the entire network of relationships with customers, suppliers and partners [6]. New evaluation frameworks include "network effect" metrics that quantify these systemic benefits often ignored in traditional analyses.

Communicating value: from analysis to strategic storytelling

Market leaders have definitely abandoned the purely quantitative approach in favor of more comprehensive methodologies that integrate:

  • Digital twins for impactful simulations: According to the Gartner Future of AI Investment Report 2025, advanced models that simulate the value of AI through digital twins in the organization are adopted by 83 percent of Fortune 500 companies [7]
  • Predictive benchmarking: The Boston Consulting Group has documented how real-time assessments are redefining the competitive landscape in technology-intensive industries [8]
  • Mapping emerging opportunities: PwC Strategy& data show a direct correlation between early identification of AI-enabled opportunities and sustained growth [9]

"Companies that rely only on traditional ROI analysis no longer even see the tip of the iceberg of the value of AI," says Dr. Marcus Lee, CTO of Novartis Digital, authoritatively. "We are seeing a complete redefinition of entire industries driven by organizations that have adopted more sophisticated evaluation frameworks." [10]

Definitely overcoming the paradox of implementation

The paradox persists but has been redefined: to gain support for ambitious AI initiatives, a compelling business case is still required, but the most transformative benefits continue to fully manifest only after implementation. The Bain & Company study "AI Value Realization 2025" documents how cutting-edge organizations have evolved the structured portfolio approach [11]:

  • Projects with quantifiable ROI: AI initiatives with immediate benefits that build momentum and confidence (40% of portfolio)
  • Transformative strategic investments: Projects with disruptive potential evaluated through broader metrics (40% of portfolio)
  • AI-driven explorations itself: AI is used to identify and evaluate new implementation opportunities, creating a virtuous cycle of innovation (20 percent of the portfolio)

The temporal dimension: thinking in cycles of transformation

The benefits of AI now manifest themselves in interconnected transformative cycles, rather than in linear steps, as highlighted by the IBM Institute for Business Value report "AI Transformation Cycles" (March 2025) [12]:

  • Optimization cycle (0-12 months): Operational improvements that create the foundation for deeper transformations
  • Reinvention cycle (1-2 years): Redefining decision-making processes and operating models
  • Disruption cycle (2+ years): Business model transformation and creation of new market paradigms

Maturity in AI adoption in 2025 is measured by the ability to manage these three cycles simultaneously, rather than progressing linearly from one to the other.

Conclusion: The future belongs to pragmatic visionaries

The organizations that are dominating AI adoption in 2025 are not simply those with the most advanced technologies, but those that have developed superior capabilities for strategic investment orchestration.

The debate is no longer between financial metrics and strategic considerations, but between organizations that have developed integrated evaluation frameworks and those that are rapidly losing competitive relevance.

This approach requires a new kind of leadership: the ability to balance analytical rigor and transformative vision, systematic thinking and decision agility, focus on immediate results and long-term planning.

As Prof. Erik Brynjolfsson recently observed at the MIT AI Summit 2025: "AI is no longer just a tool to be evaluated, but a strategic partner in redefining the very future of the organization. Our evaluation methodologies must evolve accordingly." [13]

The profile of the winners in the era of AI 2.0 is now clear: these are the organizations that have developed the ability to evaluate technology investments not just as costs and benefits, but as catalysts for transformation in an ever-evolving digital ecosystem.

Sources:

[1] World Economic Forum, "AI Investment Strategies Panel," Davos 2025, January 2025.
[2] Kahneman, D., et al, "Temporal Discounting in Corporate AI Investments," Harvard Business Review, March 2025.
[3] McKinsey Global Institute, "AI Investment Strategies 2025," April 2025.
[4] MIT Technology Review, "The New Era of Human-AI Collaboration," February 2025.
[5] Deloitte, "AI Resilience Report 2025," March 2025.
[6] Forrester Research, "AI-Driven Business Ecosystems," April 2025.
[7] Gartner, "Future of AI Investment Report 2025," March 2025.
[8] Boston Consulting Group, "Competitive Advantage in the Age of AI 2.0," February 2025.
[9] PwC Strategy&, "Early AI Opportunity Identification and Market Growth," January 2025.
[10] Lee, M., "Beyond Optimization: AI as Strategic Partner," Digital Pharma Summit, March 2025.
[11] Bain & Company, "AI Value Realization 2025," April 2025.
[12] IBM Institute for Business Value, "AI Transformation Cycles," March 2025.
[13] Brynjolfsson, E., "AI as Strategic Partner," MIT AI Summit, April 2025.

Resources for business growth

November 9, 2025

Regulating what is not created: does Europe risk technological irrelevance?

Europe attracts only one-tenth of global investment in artificial intelligence but claims to dictate global rules. This is the "Brussels Effect"-imposing regulations on a planetary scale through market power without driving innovation. The AI Act goes into effect on a staggered timetable until 2027, but multinational tech companies respond with creative evasion strategies: invoking trade secrets to avoid revealing training data, producing technically compliant but incomprehensible summaries, using self-assessment to downgrade systems from "high risk" to "minimal risk," forum shopping by choosing member states with less stringent controls. The extraterritorial copyright paradox: EU demands that OpenAI comply with European laws even for training outside Europe-principle never before seen in international law. The "dual model" emerges: limited European versions vs. advanced global versions of the same AI products. Real risk: Europe becomes "digital fortress" isolated from global innovation, with European citizens accessing inferior technologies. The Court of Justice in the credit scoring case has already rejected the "trade secrets" defense, but interpretive uncertainty remains huge-what exactly does "sufficiently detailed summary" mean? No one knows. Final unresolved question: is the EU creating an ethical third way between U.S. capitalism and Chinese state control, or simply exporting bureaucracy to an industry where it does not compete? For now: world leader in AI regulation, marginal in its development. Vaste program.
November 9, 2025

Outliers: Where Data Science Meets Success Stories.

Data science has turned the paradigm on its head: outliers are no longer "errors to be eliminated" but valuable information to be understood. A single outlier can completely distort a linear regression model-change the slope from 2 to 10-but eliminating it could mean losing the most important signal in the dataset. Machine learning introduces sophisticated tools: Isolation Forest isolates outliers by building random decision trees, Local Outlier Factor analyzes local density, Autoencoders reconstruct normal data and report what they cannot reproduce. There are global outliers (temperature -10°C in tropics), contextual outliers (spending €1,000 in poor neighborhood), collective outliers (synchronized spikes traffic network indicating attack). Parallel with Gladwell: the "10,000 hour rule" is disputed-Paul McCartney dixit "many bands have done 10,000 hours in Hamburg without success, theory not infallible." Asian math success is not genetic but cultural: Chinese number system more intuitive, rice cultivation requires constant improvement vs Western agriculture territorial expansion. Real applications: UK banks recover 18% potential losses via real-time anomaly detection, manufacturing detects microscopic defects that human inspection would miss, healthcare valid clinical trials data with 85%+ sensitivity anomaly detection. Final lesson: as data science moves from eliminating outliers to understanding them, we must see unconventional careers not as anomalies to be corrected but as valuable trajectories to be studied.