Article · Management practice

Write down the assumptions behind a forecast

A forecast is easier to update when its drivers are visible.

Rollin & Co · 20 October 2026 3 min read

Singapore skyline overlooking the water

Start with the drivers

Identify what changes revenue, staffing, purchasing and cash collection. Use assumptions that can be explained and updated. Avoid entering a growth percentage simply because the spreadsheet requires one.

Test a different outcome

Consider slower collections, a delayed contract or higher input costs. Change the relevant drivers and observe the cash effect. Scenarios help a team discuss options; they are not predictions with guaranteed outcomes.

Learn from the difference

Compare actual results with the assumptions, not just the final total. Update the drivers where evidence has changed and retain the previous version so the team can understand what was revised.

Make the model explainable outside the spreadsheet

A forecast is a structured view of what might happen, built from assumptions that can change. Write down the main drivers in ordinary language so that someone reviewing the result can understand the basis without tracing every formula. For sales, this might include expected order volumes, prices, conversion rates and the timing of delivery. For costs, it might include staffing plans, supplier prices and capacity limits. State where each assumption came from and when it was last reviewed. Distinguish a signed order from an opportunity still being discussed, and distinguish a supplier quotation from a rough estimate. For example, a forecast of higher sales may require additional staff or inventory before the revenue arrives. If the model increases sales without those related costs and cash movements, its apparent improvement may be misleading. Check that the operational story and the financial numbers agree. A simple model with visible assumptions is often easier to challenge and maintain than a complex one whose drivers are hidden.

Change scenarios deliberately and preserve the record

Keep a dated base case, then create clearly labelled alternatives for the uncertainties that matter to the decision. A slower collection scenario answers a different question from a lower sales scenario, so avoid mixing several changes without explaining them. Show which assumptions changed and which stayed the same. Identify any limits in the model, such as an excluded financing cost or an uncertain start date for a new operation. When actual results become available, compare them with the earlier assumptions rather than only with the latest revised forecast. That comparison reveals whether the model was useful and where judgment needs improvement. Keep previous approved versions so that management can understand how expectations developed. Assign an owner to each important assumption and agree when it should be refreshed. The forecast should support a decision with a range of possible outcomes, not provide a single reassuring number whose apparent precision exceeds the quality of the evidence.

General information, not advice for a specific company. Confirm the requirements that apply to your circumstances.

Contact

Discuss your reporting or company requirements.

Tell us what you need help with and any relevant deadline.

Contact us
WhatsApp