CFO are often forced to make good decisions based on poor or missing data. Given the cacophony of data now available to them, CFOs must learn to see through the noise, taking advantage of the technology available to their teams
There are few jobs at the top of the corporate ladder that have changed more in recent years than that of the CFO.
Whereas the role used to be centred solely around a firm’s financial dealings, developments in data and AI have pushed the average CFO into a new world in which their remits have expanded to be part not only of how a firm runs its day-to-day business, but also central to its future—and, consequently, its success or lack of it.
In 2022, Accenture released a report titled The Paradox of Choice for CFOs. That report, which built on the answers from one hundred CFOs from around the world, said that the overwhelming majority of those questioned—93%—felt that their current responsibilities weighed heavier on them than they had in the past.
Likewise, nine out of ten said that they were being asked to ‘call the shots’ on business-critical decisions that impacted not just their financial departments, but the entirety of their organisations.
“In the past, CFOs were viewed as more transactional and often did not have a seat at the table,” says Russell Lester, CFO of Versapay.
“They were there to manage the cash in order to keep the business out of trouble and to help avoid risk. But we’re now seeing them being viewed as operational partners with the CEO and as strategic advisors. The reason for that is because of data. It’s that which is putting the CFO in the position of being able to see across the whole business.”
This has been a trend commented upon regularly in recent years. A 2020 report from Accenture, The CFO as the Catalyst for the Data-Driven Enterprise, stated that the demand for “rich, accurate, and timely data” was growing dramatically as organisations began to see its value across all segments of their business.
It was this data, wrote Accenture, that would allow firms to ‘innovate, compete, and improve profitability’. This has had an impact on how CFOs work and in the duties of their roles.
As RSM wrote in its 2020 The Data-Driven CFO report, “often, today’s CFO is forced to make good decisions based on poor or missing data. Hindered by unsustainable levels of cost and impairment in data, they routinely rely on their best judgement to adjust for these gaps in data and reporting.”
It added: “CFOs are focused on generating required financial reporting each period while simultaneously serving enterprise and line-of-business executives who expect to receive analytics reflecting their own specifications, such as key performance indicators, time horizons, levels of summarisation, and visual formatting requirements.”
The fallout from this has been a cause of stress.
Tom Zauli, senior vice president and general manager at SOFTRAX, says that CFOs today are buried under an avalanche of data just a short time after not having any access to real-time data at all.
“This is a multi-variate problem where a change in one data point affects the others and the overall decision process,” he says.
“CFOs must consider the situation with the economy, interest rates, the valuations for potential mergers and acquisitions, the appetite of public markets, the ability to raise cash and the associated options, the past performance of existing products in terms of rate of sale, retention rate, consumption metrics, as well as the future potential of products in the development pipeline.”
This is a lot.
“Not only must they [CFOs] have means of monitoring all of these metrics, but they must also devote the time to make sure these metrics are well understood and accurate. They then must make decisions that tie these multiple variables together, considering the complex interplay of the complete set,” Zauli adds.
Outside of the challenges arising in their own roles, CFOs often encounter issues with the use of data when it comes to knowing what is useful and what is superfluous.
Jason Dess – senior managing director, global CFO and enterprise value lead at Accenture – says there are two types of data: financial and enterprise.
He says the former, depending on the industry involved, has different required levels of rigour and transparency, particularly when it comes to auditing and risk. Its enterprise-wide equivalent has many different levels of structure, with some data arising from different social media sources being entirely unstructured.
It is a problem that leads Versapay’s Lester to say many CFOs have difficulty discerning what he calls “the single source of truth.”
The sheer amount and weight of the data available means that the conflict is often in deciding which insight of many gleaned from the same source is valid. It is a problem that was highlighted by RSM’s report which said, “no single system constitutes the complete source of financial truth.”
These problems escalate and expand when one company acquires another. Now, instead of having one firm dealing with its legacy systems and attempting to apply a solution to bring everything into one structure, there is one firm dealing with the same problem two or three times over at the same time, multiplying exponentially the work involved.
“That’s something we’ve seen in our own company,” says Lester, “because we’re an amalgamation of several firms. We’ve also seen it with our clients, particularly those that are growing.”
The problems, he goes on, tend to be centred on one side of a firm’s business. “It’s the back office and that’s powering a lot of CFOs,” he says.
“But it’s also not the first place that people focus on when integrating so you end up with these bifurcated legacy systems that do not talk to each other. The product side of the organisation often has the most-modern tools, so this means that the CFO often find themselves cobbling data together using a bunch of inefficient means.”
The modern CFO, then, has seen their role shift immeasurably in recent years because of data. To ride that change, Francis Trudeau, CFO at BrainBox AI, says that they need to look at what they have with an end goal in mind of where they want to end up.
“To get data that’s useful,” he says, “you need to first centralise it and then you need to standardise it. And all that has to be done in a way that adds value.” Centralising, Trudeau says, is about bringing everything you have together in one place.
“We have all this data coming from different sources and from different systems with different processes. You need to bring these together into one repository,” he says.
“From out of that, you can create a table and optimise that. It’s difficult to do that from different sources. It’s one of the reasons that we’ve turned to cloud computing for all of our infrastructure. That’s helped put different layers of data together to create value.”
Trudeau turns to standardisation.
“The nomenclature was more manual in recent years, but to use these protocols now and digitalise the data in a way that it can be handled and processed is a different ballgame. In our case, there are large manufacturers that own a significant part of the market and use their own nomenclature. We are agnostic, so we leverage data, and we need a convention to do that. We call it ‘mapping’.”
There are other things to consider, says Trudeau, such as governance and security. There have been many leaks of private data over the years, both of consumers and of companies. All of this has led to laws that need to be complied with.
For the CFO that is looking to overhaul, refine, and standardise their data processes through automation, Trudeau returns to the idea that they understand what they want the end product to look like.
“You need to understand the end-to-end process,” he says, “and there’s a lot of human interaction in there. Once you have that, you can start automating section by section.”
Accenture gives other advice, stating that data practises and problems should be addressed at the source, with transparency built in throughout.
RSM laid out several steps it’s report:
But, perhaps key, is software, says Zauli. “From strictly a data aggregation standpoint,” he says, “tools such as SnowFlake have entered the scene, providing significant help.”
He names other, AI-centric tools such as ChatGPT, Palantir’s Foundry and Gotham software programs, and Bard from Google. It may be a steep road ahead, though.
“It is often difficult for humans to truly comprehend an exponential,” says Zauli.
“CFOs must understand that not only is there a high rate of change in technology around this problem, but this rate of change is also increasing. On top of everything else they have going on, CFOs must at minimum keep an eye on what is available, or risk being quickly left behind.”
