Key Takeaways

  • Private equity finance teams are running into budget overruns by approving generative AI software as isolated SaaS line items like Claude or Perplexity instead of funding specific workflow outcomes.
  • Justin D'Onofrio, Managing Director at Accordion, tracks software return on invested capital by measuring exact labor hours saved, such as cutting 30 hours out of a 40-hour monthly board deck preparation cycle.
  • Devin Mathews notes that software budgeting disputes are actually value disputes in disguise: companies must define exactly what output is purchased for each dollar spent.
  • Portfolio CFOs who test generative AI on internal finance workflows first gain the benchmark data required to evaluate capital requests from marketing, operations, and IT.

The Failure of Tool-Based Line Items

Private equity portfolio companies spent the past eighteen months approving software licenses on an ad-hoc basis. A marketing lead asked for Jasper, an engineering team expensed GitHub Copilot, and an FP&A analyst put Claude on a corporate card. Finance teams treated these expenses as typical SaaS seats. That approach is falling apart as model pricing shifts and seat counts expand without matching gains in output.

“The trend has been budgeting for the tool,” D'Onofrio explains. “Hey, we're going to use Claude here. And that's the way the line item is addressed from an FP&A perspective. The change in thinking is really making sure that you're budgeting for the outcome.”

When a finance department budgets for a tool, it assumes the software itself creates value. In practice, underlying foundational models change every six months, vendors shift pricing structures, and employees abandon platforms that fail to fit their daily tasks. Budgeting for the tool leaves the CFO holding the bill for idle seats.

Building the ROIC Case on Hours Saved

To establish real financial governance, portfolio CFOs must require business units to tie software spend directly to productivity baselines. Instead of approving an enterprise license based on general promises of efficiency, the investment case must specify the target process and the expected labor reduction.

“The real way to think about it now particularly from a CFO FP&A lens is we're budgeting for an outcome,” D'Onofrio says. “And that outcome might be we expect productivity of the accounting team to improve by 20% with this use of Claude.”

Consider monthly management reporting. Building board materials and running variance analyses often eats 40 hours of an FP&A team's monthly bandwidth. If software automation reduces that workload to 10 hours, the company captures 30 hours of redeployable capacity. “If we can save 30 of those 40 hours, we can now put a clear ROIC case together on why we should invest in using AI to speed up that productivity and that output,” D'Onofrio points out.

Mathews frames the shift directly: “A budgeting question is really more of a question about value in disguise. It's not about the budget. It's about, hey, we're spending what for what.”

Why Finance Must Test Its Own Shop First

Operating partners and CFOs often struggle to evaluate AI budget requests from sales or supply chain leaders because finance lacks an internal baseline for what these tools actually achieve. D'Onofrio argues that finance leaders must run AI implementations inside their own department before judging external department requests.

“Picking a workflow within your own org as a CFO is important so that if the marketing team, the operations team, the IT team comes to you with an investment case, you have that direct experience from your own team to weigh in and gut check if that's the right approach or thinking,” D'Onofrio says.

A CFO who has measured cycle times on board decks or month-end close workflows understands prompt reliability, data hygiene issues, and real labor displacement. That direct operational knowledge prevents the finance seat from either rubber-stamping unproductive software or choking off high-return automation projects across the portfolio.

Why It Matters

Sponsor-backed companies face tight exit multiples, making headcount productivity and margin expansion the primary drivers of enterprise value. Treating artificial intelligence as an uncapped research and development expense dilutes EBITDA without guaranteeing operational leverage. By demanding concrete ROIC cases based on measurable capacity gains, buyout sponsors ensure technology spend converts directly into margin flow-through at exit.