Performance Talks Series: "We're a 1 1/2 year into our AI journey. I have very little to show for it.”

Apr 30

"We're a 1 1/2 year into our AI journey. I have very little to show for it.” is what a CFO told us in a recent meeting. And they're not alone.

Gartner says: Only 7% of CFOs report high ROI from AI in their finance functions. The rest are spending, experimenting, and hoping. The board keeps asking and demanding performance. Nobody has a good answer.

Here's what we told this CFO directly: “You don't have an AI problem. You have a measurement problem.”

Most organizations skipped this hardest step. Before the tools. Before the vendors. Before the pilots. You need one thing: a KPI baseline. A specific, owned, trackable metric that AI is supposed to move. Without it, you're not running a strategy. You're running a competition.

"Every quarter, I told my team: bring me the best AI use case." Five quarters later? Still nothing worth funding.

That's when we said: “Press pause.”

Not to kill the program. To reset the discipline. Start with the “boring” stuff that actually works like Reconciliations. Prove to the organization that AI can change how work gets done, then build from there.

We’re working with CFOs who are winning that aren't chasing the most exciting use case. We’re engineering the AI business performance with them. “What specific number are you trying to move, and by how much?” That's the difference between an AI experiment and an AI strategy.