AI for ROI

The “experimentation phase” of AI in the finance function is hitting a wall of reality. As boards demand tangible results, CFOs are moving away from speculative pilots and toward tools that deliver immediate bottom-line impact. In a recent discussion with Jason Kurtz, CEO of Basware, it became clear that the divide between AI leaders and laggards is already visible in the margins.

The Boardroom gap: Idea vs deployment

According to Basware’s research, 44% of CFOs feel intense board-level pressure to act on AI, yet they lack clear guidance on how to execute. This often leads to a cycle of “experimentation for experimentation’s sake,” which Kurtz views as a major strategic pitfall. He notes that 61% of finance leaders have rolled out AI agents as an experiment just to see what the technology could do, rather than targeting specific outcomes.

Kurtz highlights the cost of this lack of focus, citing a head of digital transformation at a major European firm who spent over €1 million on AI experiments in a single year without generating a single euro in savings. The key to success is getting resources concentrated on functions that can deliver fast ROI specifically the Office of the CFO rather than treating AI as a “magic wand” that doesn’t require strategy or governance.

Redefining the “Digital Teammate”

A significant barrier to adoption is a lack of understanding regarding the technology itself; one in four finance leaders admits they don’t fully understand what an AI agent looks like in practice. Kurtz distinguishes these agents from traditional Robotic Process Automation (RPA). While RPA follows fixed, rules-based steps acting as a digital conveyor belt AI agents function as “trusted teammates.”

These agents can provide contextual, real-time guidance and handle natural language queries. Instead of just moving data, a data agent can answer complex questions such as, “Which suppliers can give us an early payment discount this month?” or identify bottlenecks in the matching engine. This shift moves AI from a simple task executor to a strategic partner that allows human managers to focus on higher-value activities.

The Case for Accounts Payable

For CFOs seeking measurable ROI within 6 to 12 months, Kurtz points directly to Accounts Payable (AP). It offers a quick path to value because AI can accelerate automation across the entire invoice lifecycle, reducing manual intervention and allowing teams to scale without adding costs.

Furthermore, AI improves decision quality by turning fragmented data into real-time insights for cash flow management. Because AI continuously learns from rich transactional data, it shortens implementation timelines by identifying process gaps automatically. In a risk-averse

environment, the “risk-adjusted return” of AP makes it the logical starting point for any finance AI strategy.

The Headcount Myth: Replacement vs. Empowerment

The narrative that AI is coming for every job makes for sensational headlines, but Kurtz offers a contrarian perspective based on historical precedent. He compares the AI shift to the introduction of the assembly line, which transformed skilled artisans into specialized machine operators and ultimately boosted overall factory employment.

In the finance world, agents are not replacing people; they are replacing the manual, repetitive tasks that bog them down. By automating the “not fun” parts of the job like chasing invoice approvals staff are empowered to be more analytical and strategic. Kurtz argues that while some companies are cutting staff, they are often using AI as an excuse to “right-size” after over-hiring, rather than because the technology has made the humans redundant.

The 80% ROI Divide

The performance gap between companies is widening. While general ROI from AI has risen from 35% to 67% in the last year, companies using “Agentic AI” through third-party solutions are outperforming the rest with an average ROI of 80%.

Kurtz attributes this to the power of data. Basware has processed 2.5 billion invoices and €10 trillion in spend, a data set that no individual company can replicate internally. He suggests a simple rule for survival: focus on what you are best at. For a manufacturer, that means focusing on products and customers, while allowing specialized experts to manage the invoice lifecycle.

The Bottom Line: The window for aimless experimentation is closing. The most successful finance teams will be those that move past the “pilot” stage and integrate embedded, governed AI solutions that treat the technology not as a science project, but as a trusted member of the team.

Comments are closed.