CFOs have long paid close attention to their business’ data, but in the age of generative artificial intelligence, they need to focus on unstructured data — information that companies collect, but that may be tucked away in silos or in formats incompatible with other data.

That unstructured information could offer CFOs a critical edge in an era where companies are looking to utilize the possible benefits of AI. To take full advantage, finance chiefs need to start thinking differently about their data management strategies, said Ben Little, CEO of knowledge management SaaS provider Bloomfire.

“The real fundamental shift that we're seeing, particularly at the office of the CFO, [is] they're saying, ‘every piece of data needs to pull its weight,’” Little said in an interview. “It is either adding to our business objectives, or we're seeing it as…a liability.”

From cost center to benefit 

The problem of unstructured data isn’t a new one for businesses. While CFOs understand the power and importance of data, many businesses collect information that they simply funnel into storage or databases without analyzing it — a solution that is quickly losing its appeal in the AI age.

Companies are “now recognizing that may not be optimal, because we have a lot of information that is redundant, outdated, trivial, conflicting, and what it's doing is it's creating business inefficiency and business risk,” Little said.

Little has served as CEO for the Austin, Texas-based data management company for nearly three years, according to his LinkedIn profile. He serves on the executive healthcare advisory council for RiverGlade Capital and as a member of the board for Home Helpers Home Care, and started his career at home services marketplace HomeAdvisor, which merged with Angie’s List in 2017, according to reports.

Figuring out how to correctly marshal their unstructured data is especially top of mind for CFOs that need to figure out how to most effectively make use of new technologies like GenAI, CFO Dive previously reported, citing data from a Komprise survey. Fifty-seven percent of respondents identified AI prep as a top challenge when it came to their unstructured data, the survey by the data management software provider found.

While tools such as AI can more quickly scan through data and provide deeper insights, those benefits disappear if the AI models are working with redundant, outdated or incorrect information. It’s important CFOs view their data management strategies with these challenges in mind, but crucially, they also need to recast how they view data management itself from a cost center to a benefit, Little said.

“If we go into this thinking that this is going to be an expense, I think we're probably setting ourselves up, not for success,” he said. “Because my personal belief is this should be a positive ROI, so it shouldn't be, ‘how much is this costing?’ It should be, ‘how much is this making?’”

Many CFOs are also wary of shelling out funds for solutions that might simply add to their costs, rather than solve their challenges: AI is likely to increase the amount of technical debt many companies have, with more than 50% of technology companies expecting that debt to reach a “moderate or severe” level by 2025, a recent Forrester report found. 

“They're very weary of, ‘oh, do this, and you're going to get this fictitious productivity gain,’” Little said of how CFOs are thinking of emerging technology and software products.

Tailoring data goals   

Another challenge is that many businesses often develop their processes in unstructured ways, as well as storing unstructured data in silos, Little said.

“Sometimes it is the lack of a system that creates four additional steps,” he said.

However, CFOs also shouldn’t look to do too much, too quickly: to best take advantage of new technologies like AI that can reduce manual tasks, complicated processes and better utilize unstructured data, “step one is specificity,” Little said. “Oftentimes I have people say, ‘oh, let's go do everything. The proverbial term [is], we can boil the ocean with this. And I say, yes, but let's go focus on something specific.”

A business leader asking for something such as a broad 20% reduction of all costs, for example, “that's to me, that's a red flag, as opposed to when someone says, ‘here's my challenge, I've noticed that my NPS has been decreasing,’” he said.

Working to better organize and utilize unstructured data with such a goal in mind, “we can positively impact average handle time, or average hold time,” he said. “We can measure a decrease in callbacks and escalations. That's a quantifiable metric that that a CFO can latch on to.”