06-reference

cfo secrets building fpa series i

2026-07-04·reference·source: CFO Secrets·by The Secret CFO

CFO Secrets — Building FP&A Part I: What Even Is FP&A?

Part I of a new four-part series on how to build and scale an FP&A function. The author frames this as a companion to last year's nine-part series on individual FP&A processes, but with a focus on how to turn the processes into a functioning machine.

⚠️ Sponsorship

Sponsored by Summation (https://www.summation.com/ai-analyst), an AI analyst tool that automates data assembly and cleanup so FP&A analysts can focus on analysis. Sponsor block is a clear paid placement with UTM tracking. No author investment relationship indicated — Summation is not in the known Campfire/Zip/Pulley/Stuut rotating pool; this appears to be a new sponsor. The sponsor's pitch (free analysts from assembly work so they can actually analyze) is editorially consistent with the issue's core thesis, though that alignment should be read as a sponsor benefit, not independent endorsement.

Summary

The issue opens with a post-merger integration story: a $1.5B acquisition with a "world-class FP&A function" turned out to be running a sophisticated system of performative accountability — budgets designed to miss slowly, explained elegantly each month, with no actual behavior change. A divisional president crystallized it: the business had inverted, serving the planning machine rather than the planning machine serving the business.

Core thesis: FP&A exists for one reason — to help the business improve its financial performance. Budgets, forecasts, dashboards, and reports have zero inherent value. Their value derives entirely from impact on actual results. The guide runner analogy: FP&A only matters if it helps the athlete run faster.

The function vs. the work distinction: Most finance leaders think of FP&A as a team (headcount, job titles). The author reframes it as an activity set — cycles and processes that create work across the entire organization, not just inside finance. That means the CFO is designing the "performance operating system of the business," not just a team or a dashboard.

The FP&A Activity Stack — 7 types:

  1. Planning (LRP, annual budget, target setting, resource allocation)
  2. Forecasting (reforecasts, scenarios, risks/opportunities, early warning)
  3. Performance reporting (monthly packs, KPIs, variance, board spine)
  4. Performance analysis (bridges, root cause, driver analysis, action tracking)
  5. Business finance (commercial/operational finance, pricing, deal models, profitability)
  6. Strategic finance (capital allocation, M&A, transformation, valuation)
  7. FP&A infrastructure (models, tools, dashboards, data definitions, automation, AI)

The "better not more" argument: The author calls out the industry's relentless push toward more — more dashboards, more forecasting, more data, more AI, more software, more consulting. Notes this cycle is playing out again with AI specifically. The counterpoint: what businesses actually need is better FP&A, not more of it.

Series roadmap:

Why this is in the vault

The "performance operating system" framing and the "better not more" counternarrative are both load-bearing concepts for RDCO's positioning. The FP&A Activity Stack is a structured vocabulary for finance-function conversations with CFO and finance buyers — directly applicable in the phData DSA role when qualifying and scoping data/AI engagements. The performative accountability story is a deployable cautionary tale for any AI deployment that adds dashboards without changing decisions.

Mapping against Ray Data Co

Strong alignment — three direct vectors:

  1. "Better not more" as RDCO's counter-positioning in AI sales. When pitching data/AI solutions at phData, the vendor-noise problem the author describes is exactly the environment DSAs operate in. Every competitor is selling "more AI" — more agents, more dashboards, more automation. The CFO Secrets framing gives a CFO-fluent vocabulary for a different pitch: FP&A that is right-sized and impact-focused, not impressive-looking. This is the positioning gap RDCO can exploit at the CFO/VP Finance buyer level.

  2. The harness-as-performance-operating-system parallel. The author's critique of FP&A that runs the business instead of serving it maps precisely to how RDCO thinks about the Claude Code harness. The harness exists to improve founder output; it is not a goal in itself. When the harness starts generating ritual work (compaction ceremonies, approval loops, redundant verification passes) that doesn't change actual outcomes, that is the same inversion Krishna was describing. This issue is a useful diagnostic prompt for harness audit cycles.

  3. FP&A Activity Stack as a client scoping tool. The seven-bucket taxonomy is a structured way to map where a prospective client's data/AI work actually sits — and where the value unlock is. A client asking for "AI for finance" is often asking for FP&A infrastructure (bucket 7), but the real value may be in performance analysis (bucket 4) or business finance support (bucket 5). Having this vocabulary available in discovery conversations improves scoping precision.

One tension to flag: The author argues FP&A should exist only through measurable performance impact, and is skeptical of sophisticated-sounding processes that don't change behavior. RDCO sells data/AI tools and consulting — by the author's standard, that work needs to justify itself through impact on client financial performance, not through implementation effort or tool sophistication. That is a useful internal constraint, not a reason to avoid the space.

Series watch: Part IV (future of FP&A with AI) will likely be the most directly relevant issue for RDCO's AI positioning thesis. Flag for priority filing when it lands.

Related