When My FP&A Model Got So Big It Qualified for Its Own Zip Code

Every finance team has its Frankenstein—the spreadsheet that starts as a modest forecast and metastasizes into a corporate public works project. Mine began as a single tab. By Q3, it was a bureaucratic anthill with 184 tabs, a pivot table the size of a congressional bill, and lookup formulas twisted like the wiring in a Soviet elevator.

Opening it was like wandering into an abandoned industrial park. Rows of half-finished logic. Whole columns marked “DO NOT TOUCH” in the desperate caps lock of the already defeated. Somewhere deep in “Churn Assumptions – APAC SMB,” a SUMIF colony had achieved sentience and was probably plotting secession.

The board loved it—at first. “So comprehensive!” they said, mistaking breadth for competence, the same way Congress mistakes 1,200 pages of pork-barrel spending for legislation. Then, like a scandal breaking in slow motion, they realized it was unusable.

And that’s the rot at the core of most FP&A models: They’re not built to make decisions. They’re built to survive questions. You bury assumptions under enough tabs and cell references, and nobody has the energy to challenge them. Complexity becomes the shield. You can’t lose an argument if no one can find the math.

The cost? Real and immediate. Models this bloated turn CFOs into archivists of their own work. Decisions crawl. Risk hides in the shadows. And by the time the numbers make sense, the opportunity that demanded them is already dead.

If your model is too complicated to explain in one breath, it’s already a liability.

The Four-Step Exit From Spreadsheet Authoritarianism

1. Physics Before Cosmetics
Tie every metric to business drivers that don’t change—unit economics, sales velocity, churn patterns—rather than the political weather inside the C-suite. ARR doesn’t get a new definition every quarter.

2. The 90-Minute Rule
If you can’t refresh the forecast in an hour and a half, it’s a liability, not a tool. Keep only what moves the outcome: growth rate, churn, CAC, hiring velocity. Everything else is decoration.

3. Scenario Batching
No more “custom what-ifs” for every executive mood swing. Establish three sets—base, stretch, crisis—and teach leadership to work inside them. If they can’t, the problem isn’t your spreadsheet.

4. One-Tab Governance
One clean, human-readable summary tab. Board-ready. No scavenger hunts, no hidden macros, no “secret” tab only the model owner understands.

The Real Job of a CFO

Founders and CFOs aren’t paid to run data mausoleums. The job is to make fast, confident decisions under pressure. The tool that lets you do that is your model. The tool that keeps you from doing that is also your model—when it’s designed like a city zoning map.

A good model is a scalpel. A bad model is a museum exhibit. And nobody runs a business from a museum.

If your CEO calls at 6 a.m. asking whether you can extend runway by three months, you should be able to answer before coffee—without booting up a file that now requires its own municipal utilities department.

Because once your model needs its own zip code, it’s not a decision tool. It’s a political artifact. And political artifacts don’t run companies. You do.