Chief financial officers (CFOs) must invest in artificial intelligence (AI) within the next few years, or their companies will be left behind.

But to get the full benefits, and thus competitive advantage, from such investments CFOs must look beyond projects that only aim to modernise the function.

“There’s nothing wrong with using AI to modernise the finance function. It’s very important work,” says Clement Christensen, director in the Gartner Finance practice. “However, the most impressive rewards of AI will fall to the CFOs who think bigger about how the technology can fundamentally change the way their company does business.”

The top priorities should be to improve the organisation’s data architecture to support future AI goals; to invest in citizen data scientists so that AI production can be rapidly scaled where successful; and to redesign the organisation’s reporting suite so that is best aligned with internal customer needs rather than with traditional “finance tasks”.

“Most CFOs are aware that to reach their functional digitalization goals, they need to pursue more experimental, less familiar digital technology projects,” says Christensen. “Despite this, they follow use-case-focused approaches to AI projects that tend to have a bias towards modernising and improving familiar processes to drive easily-quantified ROI gains.”

For example, a common use case of AI is to use machine learning to predict customers prone to late payments and issue earlier payment reminders to such customers or chase late payers automatically. Reducing late payments has a clear ROI in that it will improve a company’s cash flow.

However, this is not using AI to do anything new. Finance would have chased those payments anyway, albeit less efficiently.

On the other hand, something truly transformative could be using AI to identify likely late payers at the sales stage, so that sales prospects are prioritised according to which is likely to pay most promptly.

This has the potential to transform a business’ approach to mitigating late payments and improve cash flow even further while reducing the need to chase for payments in future, freeing up finance function time for higher-value work.

“The transformative option doesn’t have an immediately measurable ROI like the modernisation option, but the ultimate payoff is potentially much bigger,” says Christensen.

To avoid the inherent bias of a use-case-focused approach to AI projects, Gartner experts advise CFOs to start with a problem that needs solving, not a process that needs modernising.

“Just a small shift in mindset in how CFOs think about deploying AI can make that the difference between a project that modernises a business and a project that transforms it, which is where the greatest competitive advantages lie,” says Christensen.