The Source for CFO Thought Leadership

Ep 49: AI’s Early Returns

Ep 49: AI’s Early Returns

In this Planning Aces episode, Jack Sweeney and thought leader Brett Knowles explore FP&A lessons from three CFOs tackling AI’s promise and limits. Craig Foster (Pax8) sees AI as an efficiency engine, projecting “20% more with 20% less” via agentic tools across functions and for SMBs. David Obstler (Datadog) outlines monetization—AI-native demand, LLM-powered products, and internal coding and finance agents—while prudently building ROI evidence. Ben Gammell (Brex) tempers hype, saying humans still forecast better beyond 90 days, yet crediting AI with faster analysis. Together, the trio charts a spectrum from pragmatic efficiency to skepticism.

Brett Knowles – Key Takeaways

Brett Knowles spotlights the spectrum of AI maturity: from skepticism about forecasting to aggressive deployment across products and functions. He argues ROI is the wrong early question; focus on cost discipline, leading indicators, and avoiding incremental spend. AI should augment teams—“another seat at the table”—while humans stay in the loop. He expects productivity: onboarding cycles collapsing, agents boosting SMB agility. He notes CFOs are candid about efficiency tradeoffs as AI embeds across workflows.

Our Planning Aces: Where They Differ

  • Primary value lens:
    • Craig Foster (Pax8): Operational efficiency & SMB enablement—“20% more with 20% less,” agentic tools across functions and partners.
    • David Obstler (Datadog): Platform monetization & product coverage—AI-native demand, LLM-powered features, internal agents; prudent, staged ROI proof.
    • Ben Gammell (Brex): Analytical acceleration, human-led forecasting—AI excels at diagnostics; beyond ~90 days, expert intuition still edges models.
  • Forecasting stance:
    • Skeptical: Gammell—models miss strategy shifts, ramp dynamics, new markets.
    • Execution-forward: Foster—focus on throughput, cycle times, headcount leverage now.
    • Enable-then-measure: Obstler—build capability, track adoption/revenue, prove ROI over 2–3 years.
  • Time horizon & scaling:
    • Near-term lift: Foster (immediate efficiency, onboarding compression).
    • Balanced horizon: Obstler (near-term monetization + medium-term ROI instrumentation).
    • Guardrailed near-term: Gammell (use AI for speed/insight; keep humans steering forecasts).