A forecast is not there to be accurate. It is there to be useful.

That statement can sound careless until you consider the alternative: a model filled with precise numbers that leadership quietly knows are based on hope.

False precision is dangerous because it makes uncertainty look resolved. A receipt appears on a particular day. A hire begins on schedule. A customer renews at the expected amount. A project closes at the assumed margin. The spreadsheet balances, so the plan feels controlled.

The business is not controlled. The uncertainty has only been hidden inside exact-looking cells.

Honest forecasting begins with behavior

A cash forecast should reflect how money actually moves.

Collections should be tied to real payment behavior and the age of receivables, not only to invoice dates. Payroll should be mapped to pay dates along with taxes, benefits, and planned hires. Disbursements should include debt service, vendors, tools, contracts, and recurring commitments.

Then the company should add the events that are predictable but often treated as surprises: annual insurance, tax estimates, renewals, seasonality, capital purchases, and customer concentration.

None of those inputs becomes certain because it enters a model. They become visible.

Visibility is enough to improve a decision if the model makes the uncertainty explicit.

Use ranges where the business contains ranges

Some inputs are genuinely fixed. A scheduled debt payment is different from a sales opportunity. A signed contract with established payment behavior is different from an unsigned proposal. The forecast should preserve those differences.

Leadership can use a base case, a downside case, and a defined set of triggers instead of pretending one outcome deserves every decimal place.

For example:

  • committed cash outflows can be entered at their known timing;

  • customer receipts can reflect payment history and concentration risk;

  • pipeline can be separated by stage instead of multiplied into one confident total;

  • hiring can be modeled with a range of realistic start dates;

  • discretionary spending can be labeled so leadership knows what can move.

The result may look less tidy. It is more useful because the team can see which assumptions matter.

Variance is information, not embarrassment

Teams sometimes defend a forecast because missing it feels like failure. That behavior destroys the model’s value.

When the actual result differs from the forecast, the first question should not be who made the wrong prediction. It should be what the variance teaches the company.

Was collection behavior slower than the team assumed? Did a pricing change affect close rate? Did delivery cost move with a different customer mix? Did the hiring timeline reveal a capacity problem? Did a supposedly fixed cost turn out to be variable?

The answer should update the next forecast and, where necessary, the operating decision.

A forecast becomes stronger through repeated contact with reality. Protecting the old version defeats the purpose.

The weekly update is part of the system

A 13-week cash forecast is useful because it creates a recurring operating rhythm. Each week, leadership can update receipts, commitments, risks, and choices while there is still time to act.

The cadence should answer:

1. What changed since the last review? 2. Which assumption created the change? 3. What is now at risk? 4. Which decision should move earlier? 5. Who owns the next action?

The model is not finished when the spreadsheet is built. It becomes a management system when the answers affect collections, payables, spend approval, hiring, financing, and investment.

Precision should follow evidence

There is a place for exact numbers. Use them where the commitment is exact. Use ranges, probabilities, and scenarios where the evidence is uncertain. Do not make the entire forecast vague, and do not make the uncertain parts appear fixed.

This is the standard I prefer: committed items are explicit, assumptions are owned, uncertainty is visible, and the model is updated often enough to create early choices.

The goal is not to predict every week perfectly. The goal is to avoid being cornered by an outcome the business could have seen and addressed sooner.

False precision makes a forecast look finished. Honest forecasting keeps it useful.

To invite Arron into a leadership conversation about forecasting and decision design, visit the speaking page. Explore related podcast and video discussions in the media library.