Because what didn’t happen?

We didn’t see the customer flight risk.
We didn’t reroute spending early enough to seize a growth opportunity.
We didn’t challenge the sales assumptions that were propped up by pure hope.

But hey, I nailed the margin.
That’s what we were told to care about. Forecast accuracy. Within 3%. Within 5%. Whatever the goal was that quarter.
Just be “close.”

And it hit me: we were optimizing for the wrong thing.
We were running a casino, not a company.
It didn’t matter why we were right, just that the numbers matched.
But being right for the wrong reasons? That’s not finance. That’s luck.

And the house always loses eventually.

Somewhere along the way, finance got addicted to precision.
We trained our teams to sweat the decimals.
To treat the forecast as a finish line instead of a flashlight.

And I get it—forecast accuracy sounds virtuous.
Discipline. Rigor. Control.
It makes us look competent in board decks.

But here’s the thing nobody wants to admit out loud:

Forecast accuracy is often a lie we tell ourselves to feel safe.
Because if the number’s close, it means we’re in control.
If the model’s tight, it means we understand the business.
If variance is low, then we’re good at our jobs.

Except… no.
Most “accurate” forecasts are just artful manipulations of last month’s reality.
They’re momentum maps.
They don’t reflect what’s coming.
They reflect what we’re too scared to change.

You know what I’d rather see?
A wildly “inaccurate” forecast that started a real conversation.

One that made someone in product say: “Wait, what’s driving that dip?”
Or got sales to finally admit that the Q3 pipeline is mostly fluff.
Or forced leadership to cut bait early instead of dragging a bad plan into Q4.

Because the point of forecasting isn’t to be right.
It’s to be useful.

Useful to decision-making.
Useful to timing.
Useful to spotting red flags before they’re waving in our faces.

We’ve built entire careers around polishing predictions no one’s brave enough to act on.
But finance isn’t supposed to be an art gallery.
It’s supposed to be a pressure test.
And if no one’s sweating, we’re not pushing hard enough.

I had a VP once who said, “I don’t care if your model is beautiful. I want it to piss someone off.”

He wasn’t being a jerk.
He was asking us to break the spell.
To stop sugarcoating every projection with caveats and context that let the business off the hook.

He wanted us to model what might happen—not just what was safe to show.

And that meant being wrong.
A lot.
But wrong in ways that revealed blind spots.
Wrong in ways that forced people to defend their bets.
Wrong in ways that actually got us closer to the truth.

Because the real measure of a forecast isn’t how well it matches actuals.
It’s how much action it drives.

Let me tell you a story.

We were working with a hypergrowth startup—classic case: big valuation, no infrastructure, three people pretending to be twenty.
The board wanted forecast accuracy tightened to under 2% variance.
Sure. Let me just hire God real quick and I’ll make that happen.

Instead, I blew the forecast up.
I layered in risk ranges. Gave them probabilistic scenarios. Showed them three paths and the assumptions that would tank each one.
It was messy.
Not “accurate.”
But you know what happened?

They pulled forward hiring for a key role that ended up saving their largest customer.
They delayed a major investment that looked good on paper but was built on bad CAC data.
They ran leaner and grew smarter.

Variance was all over the place.
But they made better calls.
And finance finally had a seat at the adult table—not as the compliance cop, but as the co-pilot.

The obsession with forecast accuracy has a body count.
Not literal, but reputational.

Finance teams burnt out trying to thread needles that didn’t need threading.
Business units avoiding us because they’re tired of “gotcha” variance calls.
Analysts spending hours reconciling why something was off by 1.7% instead of asking, “Should we have even expected this to happen?”

We’re teaching people to fear being wrong instead of rewarding them for seeing around corners.
We’re incentivizing polish over insight.
And worst of all, we’re making finance boring.
Predictable. Safe. Politely useless.

So what do we do?

We change the question.
Instead of: “How close are we?”
We start asking: “What could this trigger?”

A forecast isn’t a test. It’s a prompt.

Does it force a decision?
Does it challenge an assumption?
Does it create a sense of urgency?

If not, throw it out.
I don’t care if it’s “right.”

Build models that are designed to age fast.
That flex with inputs, not fight them.
That show cracks early so the business doesn’t fall through them later.

And teach your team to defend forecasts not like a courtroom brief, but like a working hypothesis.

What do we know?
What are we guessing?
Where do we expect to be surprised?

Build that into the meeting. Make it part of the culture.
Forecasting becomes fun when it becomes dangerous.
Not in a reckless way—but in a wake-you-up way.

We’re not here to tell bedtime stories with numbers.
We’re here to tell the truth—even if it changes tomorrow.

There’s something beautiful about letting go of the god complex.

Finance doesn’t have to be the department that “knows.”
We can be the ones who ask the hardest questions.

The ones who make the risk visible.
Who surface the tradeoffs.
Who raise their hand and say, “Hey, I don’t think this adds up—and I think that matters.”

We become dangerous when we’re honest.
Not when we’re perfect.

And that honesty doesn’t live in the “variance to plan.”
It lives in the friction. The dialogue. The recalibration.

You want to fix forecasting?
Then stop trying to impress people with how right you were.
Start daring them to make a better call.

And if you’re still chasing 2% variance like it means something…

…you might just be playing the world’s most expensive game of darts. Blindfolded. In a hurricane.

I’d rather be the one who tells you the building’s on fire—even if I miss the floor number.

I used to believe that too.

Until a friend of mine who works at a 120-year-old manufacturing company told me how they completely transformed their forecasting—and turned that narrative on its head. Hearing their story taught me something about where the real forecasting value comes from—and why most companies, old or new, miss it.

This is their story. And if you’re a CFO or operator thinking your forecasting is “good enough,” I’d take a closer look.

The Setup: Complexity Hiding in Plain Sight

The company made precision-engineered components. Big industrial clients. Global supply chains. Multiple product lines.

On paper, they had forecasting “covered”:

• Monthly P&L forecasts
• Variance reports by region
• Management reporting deck

Looked fine. Except—sales kept surprising to the upside or downside. Inventory swings caught them flat-footed. Cash flow forecasts were off by 10-15% regularly.

The board was asking questions. Operators were frustrated. Finance was tired.

That’s when my friend’s team decided to change things.

The Problem: The Forecast Was Too Pretty

Here’s what they found:

• The forecast was driven by a single, consolidated model—beautifully formatted.
• Inputs came from high-level rollups—often averages of averages (aka spreadsheet fantasy math).
• There was little input from actual operators.
• Scenario planning? Nonexistent.

In short, the forecast was too pretty. It smoothed over complexity instead of surfacing it.

You could almost hear the board collectively nodding—right up until the numbers blew up.

Sound familiar?

The Shift: Building Forecasting Value from the Ground Up

They didn’t overhaul everything overnight. They started with mindset shifts—then tactical changes.

1. Reframe Forecasting as an Operating Tool

First, they had to stop treating forecasting as a Finance-owned report. They reframed it:

Forecasting = Operating Decision Support

That meant operators had to own inputs. And Finance had to facilitate, not dictate.

Or as my friend put it: “We stopped being the spreadsheet police and started being copilots.”

2. De-layer the Model

They decomposed the monolithic model:

• Product-level drivers for Sales
• SKU-level inventory forecasts
• Region-specific FX and COGS assumptions
• Cash forecasting tied to actual receivables/payables behavior

Was it messier? Yes. Was it more accurate? Absolutely.

And bonus: once operators saw their own assumptions reflected, they started caring. A lot.

3. Implement Scenario Planning

They added structured scenario planning:

Now Finance and Operators had a shared language for planning.

As my friend put it: “No more deer-in-headlights in ops meetings.”

4. Tighten the Forecasting Cadence

The old cadence? Monthly, and mostly for board reporting.

They shifted to:

• Monthly formal re-forecast
• Bi-weekly business review forecasts (lighter)

Why? In volatile markets, waiting 30 days to update your view is like driving a race car while staring in the rearview mirror.

5. Align Forecasting with Business Questions

They stopped asking: “Is the forecast accurate?”

They started asking: “What decisions does this forecast inform? And is it good enough for that?”

Examples:

• Inventory build decisions for Q4?
• Hiring plans for a new production line?
• FX hedging levels for Europe?

Forecast accuracy isn’t the goal. Decision usefulness is.

The Result: Forecasting Became a Competitive Weapon

Here’s what changed in six months:

And the biggest win? They navigated a global supply chain shock far better than peers—because they had already modeled the scenario and knew where their exposure was.

Or, as my friend said after one tense board call: “We looked like we had a crystal ball. We didn’t. We had practice.”

Lessons Learned: Forecasting Value Comes from the Process, Not Just the Model

What this old company taught me:

• Forecasting value = alignment + insight + agility.
• The prettiest model in the world is useless if it’s not tied to operator reality.
• The Finance team that asks better questions wins.

And here’s a funny analogy I use with CFOs now: Your forecast isn’t a crystal ball. It’s a flight simulator. The more you train in it, the better you handle turbulence.

Why This Matters for CFOs and Operators

Too many companies think they’ve “checked the forecasting box.”

But here’s the test:

• Is your forecast built with operator input?
• Does it inform key operating decisions?
• Does it adapt as reality changes?
• Can your team run scenarios fast when needed?

If the answer isn’t a clear yes—you’re leaving value on the table. And in volatile markets, that’s a dangerous place to be.

Forecasting Is a Muscle You Build

This article took real time to write because I want more CFOs and operators to see forecasting not as an obligation, but as a competitive edge.

If you found value in it, please share.

And if you want to go deeper—whether it’s redesigning your forecasting process, building smarter models, or up-leveling your Finance team’s decision support game—I offer 1:1 consulting for Finance pros ready to level up. DM me if you want to talk.

And I’ll leave you with this question: If a board member asked tomorrow, “What’s the scenario plan if X happens?”—how fast could your team answer?

If that question makes you sweat—it’s time to fix it.