AI may be dominating boardroom agendas, but its financial impact is proving harder to pin down. Nearly eight in ten companies report using generative AI, yet just as many report no significant bottom line impact, according to research published by McKinsey. Another study by the Massachusetts Institute of Technology (MIT) found that 95% of organisations that have integrated AI into their operations have seen zero return.
For CFOs, this creates a growing tension: how do you reconcile modest, sometimes disappointing, early results with the sweeping predictions about what today’s systems are supposed to become at scale? Many organisations are grappling with a familiar pattern: pilot projects that never graduate, promised savings that never materialise and dashboards that look impressive but fail to influence a single decision.
Are the tools still too immature to deliver? Or are early reports missing the quiet ways employees are already using AI to boost productivity? Here, three CFOs share how they are navigating this messy middle – and what a realistic path to AI ROI looks like.
Forget the bottom line for now. This is a time for experimentation and discovery, not immediate efficiency gains. Chasing quick ROI too early risks stifling innovation before it has a chance to bear fruit. My priority is on the insights we are uncovering at Tata and the innovations we are creating. At this stage, it’s the knowledge and experience we build with AI that will become our long-term competitive advantage.
I try to resist short-term pressures from investors and analysts to deliver instant results. It is important to accept that certain AI initiatives may show a negative ROI today, particularly when developing models for complex corporate tasks. But this is a strategic trade-off: once these models are fully trained and deployed at scale, they can deliver a significant competitive advantage
Investing in AI is no longer optional. If competitors, talent or vendors embrace it, companies that fail to keep pace risk being left behind. Neglecting technology adoption can also make it harder to attract and retain top talent, as employees naturally gravitate toward organisations that empower them with modern tools.
For me, AI ROI isn’t just about immediate financial returns, it’s a strategic investment in future capabilities and ecosystem relevance. For CFOs, the focus today should be on learning, experimentation and boosting individual productivity, even if the measured ROI appears negative in this early R&D stage.
Protect experimentation budgets, educate boards and investors and frame AI as a multi-year capability build that will determine long-term competitiveness.
The value of AI lies in its measurable impact across three connected dimensions: efficiency gains, revenue generation and effective risk-management. Efficiency tells you where AI can relieve pressure on the organisation today. Revenue impact shows whether those insights are actually shaping commercial outcomes. And risk management ensures the entire effort is scalable, compliant and grounded in reality rather than enthusiasm.
AI’s clearest ROI often comes from internal efficiencies; automating routine tasks and speeding up core processes, such as reducing month-end close times from 30 days to 10. Hard metrics, such as cutting the month-end close from over thirty days to ten, help quantify these gains.
Where I see huge potential for AI to become a genuine revenue stream is in our customer service ticketing platform. We are a subscription business so if there’s a way to improve return processes by collecting customer feedback more smoothly that could have a real impact on the bottom line.
To make AI work, businesses must be prepared to invest in dedicated internal resources. At Cymbiotika, this meant hiring two specialists focused solely on identifying and implementing AI opportunities across the organisation. This ensures that AI investments aren’t wasted, but guided by dedicated exploration and a clear strategy.
Finally, companies shouldn’t overlook AI features already built into their existing systems, which can deliver faster results without additional spending. By leveraging integrated solutions, such as bill capture or supply chain modules, I’ve accelerated results while avoiding unnecessary spending.
At Pet Lab, our emphasis is on making existing operations more effective and scalable using AI, rather than generating new revenue streams immediately.
Take transaction reconciliation. We process hundreds of thousands of transactions every year. But we’ve automated 95% of this work. What was once entirely manual now requires only a small check. This not only frees my team from tedious, yet critical, tasks, but also allows them to upskill and focus on higher value work. It’s a win for retention, too – people want meaningful work, not endless spreadsheets.
AI is also transforming how we manage inventory. Getting the right stock into the right warehouse at the right moment is critical to meeting customer demand and AI dramatically improves accuracy, reducing the risk of overstocking. While this doesn’t directly generate revenue, it strengthens our ability to deliver during critical periods. That’s good for reputation and customer retention.
Looking ahead, the biggest opportunity lies in predictability. We’re using AI to forecast demand patterns, particularly around major events such as Black Friday, Halloween and Christmas. By anticipating which products will be needed where, we can align our supply chain perfectly with peak demand, maximising sales without adding unnecessary risk or cost.
This transformation introduces significant challenges in AI explainability, data privacy and algorithmic bias that finance leaders must confront. Ultimately, humans remain the ones making the final decisions. While AI can identify patterns and accelerate processes, it cannot grasp the complete context the way a person can.
While AI may not yet be a direct revenue driver, it’s an incredibly powerful tool for making the revenue we already generate work harder and smarter.
AI may be dominating boardroom agendas, but its financial impact is proving harder to pin down. Nearly eight in ten companies report using generative AI, yet just as many report no significant bottom line impact, according to research published by McKinsey. Another study by the Massachusetts Institute of Technology (MIT) found that 95% of organisations that have integrated AI into their operations have seen zero return.
For CFOs, this creates a growing tension: how do you reconcile modest, sometimes disappointing, early results with the sweeping predictions about what today’s systems are supposed to become at scale? Many organisations are grappling with a familiar pattern: pilot projects that never graduate, promised savings that never materialise and dashboards that look impressive but fail to influence a single decision.
Are the tools still too immature to deliver? Or are early reports missing the quiet ways employees are already using AI to boost productivity? Here, three CFOs share how they are navigating this messy middle – and what a realistic path to AI ROI looks like.
