Whenever I see a number that tidy, my journalistic spidey-sense starts tingling. These figures often come from studies sponsored by the very companies selling the shovels in the gold rush. In this case, the research comes via Basware, a major player in automating finance, in a report with Financial Times Longitude. That doesn’t make the number false, but it does mean we need to look past the headline. So, let’s unpack the reality behind AI ROI finance and see if this is a genuine signal of a fintech transformation or just some very clever marketing.
Deconstructing the 136% Promise
First, let’s give the number its proper context. The study published on Financial IT highlights that this 136% ROI isn’t an overnight jackpot. It’s a return calculated over a three-year period. This is a crucial distinction. We’re not talking about a slot machine paying out instantly; this is a long-term strategic investment, and understanding that is the first step in any credible AI cost-benefit analysis.
The report suggests that 82% of businesses investing in AI are seeing revenue increases. That’s a compelling figure. So where is this value actually coming from? Is some generative AI in the basement magically inventing new business models? Not quite. The truth is far less glamorous, but arguably much more powerful. The gains are rooted in efficiency and automation savings.
The core of the value proposition from companies like Basware is about tackling the mind-numbing, repetitive, and expensive manual processes that clog up the arteries of most large organisations. Think about processing invoices. It’s a critical function, but it’s also a perfect storm of manual data entry, cross-referencing, and approvals—all of which are prone to human error and deliberate foot-dragging. This is the unsexy, high-volume work where AI, specifically machine learning, truly shines.
The Tedium is the Point
This isn’t about asking an AI to write a sonnet about your quarterly earnings. It’s about giving it a mountain of invoices and telling it to sort, verify, and process them with superhuman speed and accuracy. Basware’s InvoiceAI, for instance, was trained on a dataset of 2.3 billion invoices. That’s a scale of experience a human team could never hope to achieve.
Let’s use an analogy. Imagine your finance department is a massive, old-fashioned library. Previously, to find a single piece of information, a legion of librarians had to manually search through dusty tomes, cross-reference card catalogues, and run up and down ladders. It worked, but it was slow and horribly inefficient. AI is like a digital search engine for that entire library. It has already read every book, knows where every piece of information is, and can retrieve it instantly.
What happens to the librarians? According to the study, 75% of them are “refocused on strategic work”. Instead of just finding the books, they’re now analysing the information within them to find patterns, guide strategy, and add genuine value. They’re no longer just administrators; they’re analysts. This is the promise of automation savings: not just cutting costs, but elevating human capital to do what humans do best—think critically.
The CFO’s Dilemma: Getting the Cheque Signed
Of course, this all sounds great on paper. But picture the Chief Financial Officer. Their desk is piled high with proposals for expensive software, all promising to save money… eventually. In a world of tight budgets, “spend money to save money” can be a very tough sell.
This is where the narrative is shifting. As Jason Kurtz, CEO of Basware, noted, prioritising AI investment “becomes essential for gaining CFO and boardroom approval.” It’s moving from a ‘nice-to-have’ technology project to a competitive necessity. The argument is no longer just about efficiency; it’s about survival. If your competitors are automating their back-office and freeing up their best minds for strategic tasks while you’re still drowning in paperwork, you are actively falling behind.
Conducting a thorough AI cost-benefit analysis becomes the key to unlocking these investments. A few things to consider:
– Direct Costs: Licence fees, implementation, and integration work.
– Indirect Costs: Training staff and managing the cultural shift away from manual processes.
– Direct Savings: Reduced headcount for manual tasks, elimination of late payment fees, and capturing early payment discounts.
– Strategic Value: What is the value of faster financial closing? What is the benefit of having a real-time view of cash flow? How much is it worth to free up 75% of your finance team to focus on growth instead of admin?
This last point is the hardest to quantify but is arguably the most important element of AI ROI finance. It represents the shift from a defensive, cost-cutting mindset to an offensive, value-creating one.
It’s Not Just Plug-and-Play
Before you rush off to write a blank cheque, a dose of realism is in order. Achieving a 136% ROI is not as simple as buying a piece of software. The success of these systems hinges on one critical, often-overlooked factor: your data.
Basware’s AI works because it has been fed a clean, structured diet of billions of invoices. But what does the data inside your own organisation look like? For most, it’s a messy, siloed, inconsistent tangle. The old adage “garbage in, garbage out” has never been more relevant. An AI is a powerful engine, but if you fuel it with dirty data, it will sputter and stall. A significant part of any AI project is the unglamorous but vital work of cleaning up your data house first.
And what about that other statistic? The 75% of employees refocused on “strategic work”? My question is always: what happens to the other 25%? The press releases never seem to mention them. The uncomfortable truth of this fintech transformation is that while it creates new, higher-value roles, it inevitably makes others redundant. Navigating that human impact is a challenge that requires transparent leadership, not just clever technology.
So, is the 136% ROI figure real? Yes, I believe it is—for organisations that are prepared to do the hard work of implementation, data cleansing, and process change. This isn’t magic; it’s the result of applying immense computational power to well-defined, high-volume problems.
The real revolution isn’t happening on a flashy main stage. It’s happening quietly in the accounts payable departments of the world’s largest companies. This is the engine room of the new economy, where brutal efficiency is being forged. The question you should be asking isn’t whether you should invest in AI, but how quickly you can get your house in order to do so effectively.
Is your company truly ready to make the trade: to swap manual drudgery for a shot at that 136% return? And more to the point, what’s your plan for the people whose drudgery just got automated?
