CFOs and their fellow business leaders are continuing to experiment with artificial intelligence, but getting beyond that initial testing stage is becoming steadily more important as costs rise and the number of AI tools potentially available for use multiples. The “first mile is a very quick mile with AI,” Lilac Schoenbeck, SVP of Quadient Digital said.
Users, including CFOs, can often feel “quite dazzled by the potential of the technology” in the initial experimentation phase, Schoenbeck told CFO Dive in an interview. That’s encouraging and can prompt continued experimentation, but as any marathoner could tell you, “the first mile, in some ways, is the easiest mile [and] then the challenges become, how are we actually going to scale this across the organization?” she said.
In-house data quirks
For businesses, determining that often requires debating whether the organization should buy AI solutions, or build their own.
“At some moment, you're asking the CFO organization, ‘are you a software development shop, or is this something that actually you'd prefer to leave to the professionals?’” Schoenbeck said.
Schoenbeck has spent most of her career in the software industry, where a “consistent theme” has been working with B2B software providers selling largely into the CIO and CFO organizations, she said.
Schoenbeck joined Quadient in October 2025 as its SVP of product, where she oversees R&D, product management, product marketing and customer support, she said. Before joining the France-based Quadient, which provides business software for communications, accounts payable and receivable, she served as VP of business transformation for Rocket Software, according to her LinkedIn profile.
When it comes to the build-versus-buy debate regarding AI tools, the answers tend to be circumstantial: Buying solutions can be ideal for certain processes, but for those where the business has unique needs, building a bespoke tool can often be more effective, Schoenbeck said.
“Certain kinds of processes and certain types of data are quirky to you, and you can train and build your own AI to make sense of that data with all of the house rules that are appropriate for your organization,” she said.
Generally speaking, the bigger an organization, the more siloed pieces there tend to be floating around in the back office environment — for example, programs or solutions from acquisitions that have yet to be fully integrated, she said.
Another key factor is cost, with finance chiefs needing to weigh the rising expenses associated with AI tools against potential benefits. Something “every CFO is probably painfully aware of is the costs of this are often different than they anticipated, and being able to budget for it and understand how you're using resources here is probably critical to their value generated from it,” Schoenbeck said.
Opening the CFO to customer aperture
One of the other areas executives should be considering as AI becomes more entrenched in their business is that “CFOs are one of the primary conduits of communication to customers,” Schoenbeck said.
“We don't typically think of a CFO's office as managing a customer relationship, but to a great extent they are,” she said. “And in some organizations, depending on how they're structured, they are the most persistent part of a customer relationship.”
AI can be brought in to add some visibility into the pathway between a CFO’s remit and the end customer, according to Schoenbeck. This allows an organization to understand how “the CFO’s interactions with customers are impacting your potential for upsell, your churn, your customer sentiment,” she said.
In Schoenbeck’s view, this is “a fully unused lever, and it brings together these domains of how we communicate with customers, with how we financially operate and engage with customers, and I think there's actually gold to be had in those hills,” she said.
