Is AI Worth the Cost? ROI Insights for CEOs Targeting 2025 Growth

Author(s): Konstantin Babenko

Originally published on Towards AI.

74% of companies fail at AI ROI — discover what you can do to drive real results.

According to a current NTT Data digital business survey, nearly all companies have implemented generative AI solutions, while 83% have created expert or advanced teams for the technology. The Global GenAI Report, spanning respondents within 34 countries and 12 industries, showed that 97% of CEOs expect a material change from generative AI adoption. The same report states that knowledge management, service recommendation, quality assurance, and research and development are the most valuable areas for implementing generative AI.

These findings present how generative AI is perceived in a collective sense as the enabler for change. Carlos Galve, co-head of the global generative AI office at NTT Data said: “I cannot find any other technology more transformational than generative AI. It can be applied to all industries, to all value chains, to all the specific tasks we do. Give me any kind of task and I will find some part of it where generative AI will add value.”

Having put a lot of effort into building their AI capabilities, recruiting AI talent, and experimenting with AI pilots, today’s CEOs expect ROI from the innovation. Nevertheless, the full realization of AI’s potential still presents a challenge. Current research shows that only 26% of companies are equipped with the relevant capabilities to convert AI from proof of concept into value creation (Boston Consulting Group, 2024).

This article focuses on the current AI implementation in 2024 and the future trends for 2025 based on the analysis of the latest industry research. The piece will empower CEOs and C-level executives to proactively adapt their business strategies, ensuring they stay ahead of the curve in an increasingly AI-driven marketplace.

AI Value Distribution

As per the BCG report, organizations derive as high as 60% of the generative AI value from the core business functions:

• 23% Operations
• 20% Sales and Marketing
• 13% R&D
• 38% Support functions
• 12% Customer service
• 7% IT
• 7% Procurement.

It also reveals a wide divergence between industries. Sales and marketing are reported to drive the most value from AI in software, travel and tourism, media, and telecommunications industries. Customer service appears as a prime area where the value of AI usage is tangible in the insurance and banking spheres, whereas consumer goods and retail industries are experiencing massive growth in personalization through AI.

What Separates AI Leaders from the Rest

The BCG report covers a major disconnect between AI adoption. Only 4% of companies have cutting-edge AI capabilities that provide major value and another 22% (AI leaders) are reaping big benefits from advanced strategies. On the opposite end of the spectrum, 74% of companies have not yet seen tangible benefits from AI.

According to Nicolas de Bellefonds, senior partner at BCG, “AI leaders are raising the bar with more ambitious goals.” “They focus on finding meaningful outcomes on cost and topline, and they focus on core function transformation, not diffuse productivity gains.”

Let’s take a closer look at what makes AI leaders excel:

1. Core business focus. Core processes generate 62% of leaders’ AI value, with leaders optimizing support functions to deliver a broader impact.

2. Ambitious goals. By 2027, they plan to invest twice as much in AI and workforce enablement, scale twice as many AI solutions, and generate 60% more revenue growth and 50% more cost reductions.

3. Balanced approach. Over half of leaders are using AI to transform the cost of their business and a third are using AI to generate revenue compared to their peers.

4. Strategic prioritization. Leaders focus on fewer, higher-impact opportunities to double their ROI and scale twice as many AI solutions as others.

5. People over technology. Leaders allocate 70% of resources to people and processes, thus assuring sustainable AI integration.

6. Early adoption of GenAI. Generative AI is quickly adopted by leaders emerging as a modern tool for content creation, reasoning, and system orchestration, leading the curve.

Results That Speak Volumes

Over the past 3 years, AI leaders have demonstrated 1.5x revenue growth, 1.6x shareholder returns, and 1.4x ROI, outperforming their peers. In addition to superior financial performance, they are also crushing in nonfinancial areas such as patent filings and employee satisfaction, demonstrating how their people-first, core-focused strategies are driving transformational outcomes.

Challenges Faced in the Process of AI Integration

According to the BCG report, organizations experience different issues with the implementation of AI; among them, 70% are linked to people and processes. The remaining 30% covers such categories as technology (20%) and AI algorithms (10%). The survey underlines that many companies tend to think of themselves as primarily technical organizations while the human aspect is what should not be overlooked if an enterprise wants its AI endeavors to succeed.

The Human-Centric Gap

AI integration is not just about deploying the latest technology; it is about having a workforce that is prepared to accept AI-driven changes. Lack of AI literacy, resistance to change and unclear roles in AI initiatives can often derail progress. The way leaders overcome these challenges is by investing in workforce enablement and training programs as well as building a culture in which data-backed decisions are valued.

Technology and Algorithms

On the technical side, it is difficult to integrate AI into existing systems, scale solutions across departments and keep data of the right quality. Leaders tackle these issues by strategically prioritizing a few high-value opportunities, with robust infrastructure and data governance practices.

Bridging the Gap

How well you balance the technical and human parts is key to success in AI integration. Leaders put the wheels in motion for sustainable AI adoption by placing 70% of resources in people and processes, proving that it’s not just algorithms that unlock AI’s potential, but also the technology with human capital and operational processes.

Enterprise AI Perspective for 2025

The role of AI in the enterprise environment will make further progress in 2025 as an influential element of changes in business development strategies and operational activities. Therefore, as technology advances, automation will become complementary to human talent and the way organizations manage human capital will change further. In the future, the primary competitive advantage will not lie in developing or tuning LLMs, but in their applications.

Technology complement will be one of the significant trends to be noticed in the adoption of AI because of the need to have human talent plus technology talent in an organization. Instead of outsourcing jobs to robotics, enterprises will look for tools that increase the competency and efficiency of their workers. This approach keeps the tacit knowledge of the employees within the organization as a key resource.

Data assets will remain or may even become more important as we move into 2025, as the efficiency of utilizing company-specific information will turn into a competitive advantage. Therefore, organizations need to make their data AI-prepared, which goes through several stages including cleaning, validating, structuring, and checking the ownership of the data set. AI governance software adoption will also be equally important, estimated to have four times more spending by 2030.

As the adoption of AI continues to rise, questions about its use, costs and return on investment will also increase. By 2025, a new issue will enter the picture: determining how much more it could cost to expand the use of AI and how much value organizations will be getting from these investments. Solving such issues requires finding new modern frameworks and methodologies, which will supplant already known simple KPIs, and measure customer satisfaction, decision-making, and innovation acceleration.

To sum up, the role of AI in the enterprise landscape of 2025 leads to certain challenges, such as workforce augmentation, data asset management, defining cost and ROI, and dealing with disruption.

Final Thoughts

For CEOs navigating the complexities of AI integration, the insights from this article provide a clear takeaway: AI future isn’t just about technology, it’s about leveraging the power of AI to make business value real and meaningful, aligning AI capabilities with human potential.

Looking into 2025, leaders will need to think about AI not as a standalone innovation but as an integral part of the driving force of an organization’s strategy.

There is a wide gap between the leaders and laggards in AI adoption. The difference between leaders and the rest is that they are able to prioritize high-impact opportunities, invest in workforce enablement and treat AI as a tool to drive transformation, not incremental improvement. CEOs should ask themselves:

• Are we placing bets on AI initiatives directly touching our core business functions? Leaders here get 60% of their AI value, optimizing operations, sales and marketing.
• Are we ready for AI-driven change in our workforce? To bridge the human-technology gap, resources will continue to be allocated to upskilling employees and developing a data-first culture.
• Do we have the infrastructure to scale AI solutions effectively? Robust data governance and scalable systems are important because scattered pilots won’t yield tangible value.

From my experience, enterprise AI deployments show the best results when organizations think of AI adoption as a collaboration of human expertise and technological progress. This requires CEOs to implement a long-term, strategic approach: define ambitious but achievable goals, focus on fewer, high-value AI initiatives, and create a culture open to change.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI