Scenario planning under real uncertainty: a CFO playbook

August 30, 2026 · 5 min · Free, with full transcript

Most scenario planning exercises produce tidy three-scenario decks that comfort boards and mislead decision-makers. This playbook shows how to build scenarios that actually change what your company does.

Listen to the podcast

4 min

Read the full transcript

Welcome to the Leaders Insights podcast.

Today's episode, Scenario Planning under Real Uncertainty, a CFO playbook.

The base case, the bull case, the bear case.

Every board deck this year has those three columns.

And according to today's playbook, they're actively making companies dumber.

Explain that.

They're comfort blankets.

The playbook opens on a stat that should embarrass every finance team.

A 2025 review at sites found roughly 70% of corporate scenario decks

put the base case within 5% of the sales team's original budget.

So you did all that work to confirm the number you already wanted.

That's just anchoring though.

Everyone knows to distrust the base case.

Do they?

The problem isn't that people believe it.

It's that the three scenario format smuggles in a lie.

That reality lives on a single line between good and bad.

It doesn't.

In 2022, you could have had a bear case with high inflation

and a bull case with low rates.

And the actual year gave you high inflation and resilient demand at the same time.

Neither of your three columns had that combination.

So what replaces it?

I can't hand aboard 17 scenarios.

You don't.

You start with drivers, not stories.

The playbook's move is name the two or three uncertainties

that actually swing your business and that you genuinely can't predict.

For a SaaS company, software sold by subscription,

that might be net revenue retention,

meaning how much existing customers grow or shrink,

and the cost of new customer acquisition, cross them.

Now you've got four quadrants,

and at least two of them are futures nobody in the room wants to say out loud.

Give me a real one, not a quadrant, a company.

Take what happened to Zoom coming out of the pandemic.

The scenario that killed the multiple wasn't demand falls.

It was demand holds, but growth normalizes,

and everyone re-prices you as a utility.

Revenue kept rising while the stock got cut

by more than 80% from its 2020 peak.

A tidy three scenario deck can't hold, we grow,

and the market punishes us.

A driver-based one can.

Fine, but here's the uncomfortable part.

Scenarios only matter if they change a decision.

Most don't.

They get presented and filed.

That's the whole thesis,

and it's the test I'd force on any planning team.

For each scenario, write the sentence.

In this world, we would do X differently by Y date.

If you can't finish that sentence, delete the scenario.

It's decoration.

Example of an X you'd actually pre-commit to?

Trigger-based commitments.

You don't wait for the bad scenario to arrive and then debate.

You say now,

if bookings are down 15% by end of Q1,

we freeze the two data center leases

and pull forward the pricing change.

The playbook calls these tripwires.

The point is the decision is made in a calm room in 2026,

not in a panic room in 2027.

Doesn't that just replace one false comfort with another?

You've decided in advance, so you stop thinking.

The tripwire isn't autopilot.

It's a forcing function to meet.

It says, the moment reality hits this number,

the leadership team is in a room within a week.

What kills companies isn't the wrong plan.

It's the six months of, let's wait one more quarter.

Tripwires compress that to days.

What about the numbers themselves?

Finance people love a precise probability,

30% chance of recession.

That precision is theater.

Nobody knows if it's 30 or 45.

And the third decimal makes the board feel informed

while telling them nothing.

The playbook's better habit is ranges and consequences.

If this driver moves this far, here's the cash impact,

here's the covenant we breach.

A covenant, by the way, is the promise you made

your lender, break it, and they can call the loan.

That's the number that changes behavior,

not a made up percentage.

So if I'm a CFO with a planning cycle starting Monday,

what's the one thing I do differently?

Take your current deck, find the base case, and delete it.

Physically remove the column.

Force your team to work from the drivers

and the tripwires instead.

If the plan survives without its security blanket,

it was real.

If it collapses, you just learned your planning was fiction.

Better today than in the crisis.

Done.

Wanna know where you actually stand?

Take the CFO self-assessment at MBA-training.com.

Read and go deeper