TLDR: A recent Boston Consulting Group (BCG) survey of over 280 finance executives reveals a significant gap between high expectations for AI and GenAI and actual return on investment (ROI), with a median ROI of just 10%. The study identifies four key strategies for finance leaders to achieve substantial gains: focusing on tangible value, integrating AI into broader transformation initiatives, fostering active collaboration with IT and external partners, and implementing solutions in a sequential, scalable manner. The research highlights risk management and financial forecasting as the use cases yielding the highest ROI.
Despite widespread belief in the transformative potential of Artificial Intelligence (AI) and Generative AI (GenAI) within the finance sector, a new study by Boston Consulting Group (BCG) indicates that many organizations are struggling to realize significant returns. A survey conducted in March 2025 by BCG’s Center for CFO Excellence, involving over 280 finance executives globally, found that the median ROI from AI and GenAI initiatives stands at a mere 10%, falling short of the targeted 20% for many. Alarmingly, nearly one-third of finance leaders reported limited or no gains from their investments.
The study, titled ‘How Finance Leaders Can Get ROI from AI,’ pinpoints that the challenge isn’t a lack of interest or effort, but rather an inability to achieve value realization at a reasonable cost. While 44% of surveyed teams have moved into scaled deployment and 30% anticipate transformative value by the end of 2025, many face implementation hurdles, particularly concerning compliance, regulation, and auditability. A significant 73% of executives cited four or more critical barriers to adoption, with 24% identifying seven or more.
However, approximately one in five finance functions are achieving an ROI of 20% or more, demonstrating that success is attainable through specific strategies. These high-performing teams distinguish themselves by adopting four proven tactics:
1. Relentless Focus on Value: Successful teams prioritize quick wins and allocate dedicated budgets, ensuring that projects demonstrate tangible value to secure funding. Systematic tracking of value, even through proxies like reduced resource load or improved forecasting accuracy, is crucial for steering decisions and preventing development effort from being wasted on non-impactful features.
2. Broader Transformation Perspective: Integrating AI and GenAI initiatives into a wider finance transformation agenda significantly increases the probability of success. Instead of isolated pilots, top performers stitch together connected use cases, allowing underlying investments in data and technology to stretch further. This ‘string-of-pearls’ approach enables benefits to build on each other and facilitates broader organizational redesign. For instance, one consumer goods company’s CFO redesigned their financial planning and analysis department using a driver-tree model, cutting report generation time by 50% and enabling 30% faster algorithmic forecasting.
3. Active Collaboration: Finance functions achieve higher returns by collaborating actively with IT departments and external partners. Fully staffed dedicated teams within finance, rather than part-time assignments, boost success rates by 5 percentage points. Partnering with vendors to leverage their specialized AI and GenAI expertise can also raise success odds by 5 percentage points. While 55% of solutions are built in-house on average, leveraging existing capabilities from leading ERP systems like SAP and BlackLine, which embed AI and GenAI agents, can provide sophisticated, off-the-shelf tools that address data auditability and security.
4. Targeted, Scalable Execution: The most successful finance teams avoid trying to transform everything at once. They establish small, dedicated teams to lead the AI and GenAI transformation, focusing on building what’s needed when it’s needed, ensuring each step delivers tangible returns. The emphasis shifts from showpiece pilots to the steady work of driving adoption and scaling solutions. Data transformations are also best done incrementally, letting use cases inform data quality investments rather than attempting to construct a perfect data environment upfront.
The study also identified the AI and GenAI use cases that generate the highest ROI. Surprisingly, risk management tops the list, with finance leaders leveraging the technology for fraud detection and cost reduction seeing transformative results. Financial forecasting, including cash flow modeling, sales planning, and inventory management, ranks closely behind. While many finance teams prioritize efficiency gains in transactional areas, advanced functions give greater weight to use cases that unlock business value, such as improved decision-making and faster insights.
The report also clarifies the distinct roles of traditional AI and GenAI. Traditional AI excels at deterministic processing of structured data for tasks like verifying sales or reconciling accounts. GenAI, being probabilistic and strong in language, is best suited for automating financial commentary, drafting investor communications, powering natural-language interfaces, and generating Python scripts for advanced analytics. It can also facilitate the shift of more finance activity to shared service centers by overcoming language and institutional knowledge barriers.
Also Read:
- Generative AI to Reshape, Not Erase, Banking and Finance Roles, Says Accenture
- Generative AI Poised to Revolutionize Indian Treasury Operations, Boost Efficiency
In conclusion, the BCG study underscores that achieving significant ROI from AI and GenAI in finance requires a strategic, disciplined approach focused on clear goals, effective execution, and measurable impact. Finance teams seeing the strongest results are not necessarily doing more, but rather doing what works: focusing on value, scaling with purpose, and meticulously measuring everything.
