Feb 13 2026
Artificial Intelligence

The Trouble With Measuring AI ROI

Financial leaders should consider key performance indicators such as capital efficiency, control strength, revenue quality and operational resilience.

Financial institutions need to consider alternative key performance indicators (KPIs) when making the business case for artificial intelligence because the sheer volume of tools, priorities and potential strategies makes assessing ROI daunting.

Institutions reliant solely on traditional measures such as efficiency gains, productivity aims, resource reallocations and cost savings will struggle to demonstrate AI’s value.

This challenge is borne out by the numbers: 74% of about 2,500 global tech executives report that their organizations’ AI use cases are producing business value, but only 24% are achieving ROI across multiple use cases, according to KPMG’s Global Tech Report 2026. Among those same execs, 58% acknowledged traditional ROI measures were insufficient.

“AI functions as an enterprise transformation rather than a discrete deployment, meaning value emerges unevenly across automation, adoption and reinvention phases,” says Guy Holland, global leader of the CIO Center of Excellence at KPMG International. “A more accurate ROI model blends financial impact with trust, governance, risk reduction, adoption depth and cycle time improvements — metrics that reflect the real conditions under which AI creates sustained enterprise value.”

Click the banner below to read the CDW Artificial Intelligence Research Report.

 

Linking AI KPIs to Good Governance

The good news is that some financial leaders already prioritize KPIs tying AI to capital efficiency, control strength, revenue quality and operational resilience.

“KPIs that are linked to governance and enterprisewide value realization, not just efficiency, matter just as much if not more in the broader picture of enabling sustainable value,” Holland says.

KPMG found that smaller organizations with lean governance, less complex ecosystems with fewer silos and streamlined approvals typically invested more in AI and saw 3.6 times more ROI. Early emerging technology adopters also saw more ROI because they gave themselves more time to refine their strategies. Finally, transformation-focused organizations — those expecting to spend at least 50% of their tech budgets on innovation — saw 3.2 times the ROI, likely due to a compounding effect.

WATCH: Artificial intelligence will drive efficiency for financial institutions in 2026.

Broad Efficiency Claims Fall Flat With Financial Executives

Risk mitigation and accelerated cash flow are two more AI KPIs for financial institutions to consider. Many tech leaders underestimate these factors at first, but they’re “core value drivers,” not side benefits, Holland says.

KPMG found that 69% of tech execs made security, scalability and data standardization trade-offs in an effort to expedite their organization’s digital transformation, neglecting their technical debt and talent gaps in the process. Such oversights introduce unnecessary risks that AI can help address.

The Holistic Evaluation of Language Models’ open academic framework has emerged to help organizations measure the quality, safety and bias of large language models. This, in turn, builds trust and institutional buy-in for LLMs.

“Our view is that trusted AI and strong governance enable scale, allowing AI to be embedded safely into underwriting, claims, compliance and collections — functions directly tied to capital protection and liquidity,” Holland says. “These arguments land far more effectively with financial

executives than broad efficiency claims because they map the institution’s core risk and cash dynamics.”

Patamaporn Umnahanant/Getty Images
Close

New Workspace Modernization Research from CDW

See how IT leaders are tackling workspace modernization opportunities and challenges.