How CFOs Build Forecasts Management Can Actually Use

A Forecast Is a Decision Tool, Not a Prediction Competition

I have never expected a forecast to predict the future perfectly.

That is not the standard I use.

A useful forecast should tell management:

• Where the business is currently heading
• What assumptions drive that outcome
• What could change it
• What happens under different scenarios
• What those scenarios do to cash
• What management should watch
• When management may need to act

That distinction matters.

A forecast can be wrong and still have been useful if it identified the right drivers and helped management react early.

A forecast can also land close to the final result and still have been poor if nobody understood why.

Mohamed Chaudry’s Management Forecasting Framework

My forecasting process follows seven connected steps.

1. Start With Business Drivers, Not Excel

Before opening the model, understand how the business works.

Ask what actually creates revenue.

Depending on the business, that could be:

• Customers
• Units sold
• Transactions
• Subscriptions
• Stores
• Salespeople
• Occupancy
• Projects
• Average selling price

Then identify what drives the major costs:

• Headcount
• Customer or transaction volume
• Locations
• Marketing activity
• Infrastructure
• Inventory
• Supplier pricing
• Delivery capacity

Those relationships become the architecture of the forecast.

2. Separate What Management Controls From What It Does Not

Not every forecast driver should be treated in the same way.

Management has meaningful control over some variables.

Others are largely external.

Why does this distinction matter?

Because the purpose of the forecast is partly to identify where management can act.

If sales weaken, management may not control the market environment.

It may control hiring, pricing, marketing spend, collections and other cash commitments.

3. Build an Honest Base Case

Your base case should be what you genuinely believe is most likely based on the information available today.

It is not automatically:

• The budget
• The target
• The board ambition
• The investor case
• The number required to hit a valuation
• Last year’s result plus a percentage

Targets and forecasts serve different purposes.

A target describes what management wants to achieve.

A forecast describes what management currently believes will happen.

Confusing the two destroys one of the forecast’s most useful functions: early warning.

Budget vs Forecast

You need both.

The budget tells you where you intended to go.

The forecast tells you where you currently appear to be going.

4. Build Scenarios Around the Assumptions That Matter

One forecast creates false precision.

I normally want management to understand at least three cases:

If the company has limited cash headroom, I would also consider a severe-but-plausible downside.

That case should not be a disaster fantasy.

It should answer a practical question:

Do not create scenarios by changing every line by 10%.

Change the assumptions that genuinely drive the business.

Example: A Driver-Based Revenue Forecast

5. Connect Every Scenario to Cash

This is where forecasting becomes operational.

Profit and cash do not move at the same time.

A CFO needs to understand the timing.

Ask:

• If sales slow, when does cash feel it?
• If customers pay later, how much working capital is absorbed?
• If hiring accelerates, when does payroll increase?
• If inventory rises, how much cash gets tied up?
• If capex moves forward, what happens to minimum cash?
• If funding is delayed three months, does the business still have enough runway?

A P and L forecast without a credible cash view can give management false comfort.

Example: Why Cash Changes the Decision

6. Forecast the Drivers That Move Before the Accounts

Financial statements are often lagging indicators.

By the time a problem appears in the monthly accounts, the underlying operational movement may have been visible for weeks.

Forecasting these drivers gives management more time to react.

7. Review Forecast Accuracy Without Turning It Into a Blame Exercise

Forecasts will be wrong.

The useful question is why.

When actual performance differs materially from forecast, classify the reason.

Persistent errors contain information.

If sales are missed every quarter because conversion is consistently overstated, that is no longer random forecast error.

The assumption needs to change.

What Management Should Actually Receive

Management does not need to work through 25 spreadsheet tabs.

The model can contain that detail underneath.

The management view should be concise.

I would want to see:

• Where we now expect to finish
• What changed since the previous forecast
• The three biggest drivers
• Cash and runway
• Material risks and opportunities
• Relevant scenarios
• Decisions required

How to Explain a Forecast Change

Do not simply present a new number beside the old number.

Explain the bridge.

That is what makes a forecast useful to a CEO or board.

How Often Should You Reforecast?

There is no single cadence for every business.

The right frequency depends on how quickly the business changes and how much financial headroom it has.

A stable business with strong liquidity may not need the same cadence as a scale-up with limited runway and volatile revenue.

The principle I use is:

That may mean a structured monthly forecasting cycle with more frequent cash updates where liquidity is tight.

Do not update a forecast merely to make the spreadsheet current.

Update it because your view of the future has changed.

Rolling Forecast vs Annual Forecast

An annual forecast typically looks through the remainder of a financial year.

A rolling forecast maintains a consistent forward horizon by adding another period as one closes.

For example, a rolling 12-month forecast always gives management a 12-month forward view rather than becoming progressively shorter as year-end approaches.

Rolling forecasts can be useful where management needs continuous forward visibility.

But the format is secondary.

A rolling forecast built on weak assumptions is still a weak forecast.

Common Forecasting Mistakes

Mohamed’s Forecast Pressure Test

Before presenting your next forecast, ask:

Build the Commercial Layer Behind the Forecast

Forecasting is a good example of the difference between technical finance and commercial finance.

Technical capability helps you build the model correctly.

Commercial capability helps you decide which assumptions matter, challenge what management believes, understand the cash consequence and turn the forecast into action.

That is the layer CFO Academy is designed to develop: modelling, business planning, financial communication, investment analysis, funding and the judgement to turn numbers into decisions.