Forecasting Accuracy Transformation

The Situation

A multi-entity nonprofit with a $50M+ budget had no reliable reforecast system. Revenue forecasts were routinely off by more than 10%. Board confidence was eroding. Financial planning was guesswork. The organization could not make strategic decisions without reliable revenue projections.

Leadership needed to know what revenue was actually going to come in—not what had been optimistically projected at the start of the year.

The Constraints

• No existing reforecast process or cadence—had to be built from scratch.
• The first reforecast had to be delivered while simultaneously running: an enterprise gap analysis, a monthly giving program restructure, and a 10-year data-driven analysis.
• The budget process itself was also broken and needed to be redesigned—not just the forecasting layer.
• Finance, development, and data teams had to be aligned to a shared methodology and common assumptions.

What changed

• First reforecast delivered June 2023—within 60 days of joining, while carrying three other major workstreams simultaneously.
• Q1/Q2/Q3 quarterly reforecast cycle established. Each quarter: actuals reviewed, assumptions updated, projections revised. Board received updated numbers before they had to ask.
• July 2023: entire budget process redesigned. Created zero-based budgeting system aligned to the reforecast methodology. Budget built from assumptions, not prior year actuals plus a percentage. Finance used the revised forecasts for planning and cross-entity coordination.
• Channel-level tracking infrastructure built. Each revenue channel tracked separately with its own assumptions, actuals, and variance analysis. No more blended numbers hiding channel-level problems.
• Cross-functional accountability established. Managers owned their channel projections. Variance had to be explained. Surprises stopped being surprises.

Outcomes

Board regained confidence in development leadership and strategic planning became data-driven. Finance could plan across entities using reliable development projections.

If your organization is facing this

Forecasting problems are almost never a math problem. They are a process problem. There is no cadence. There are no shared assumptions. Actuals are not being compared to projections in any systematic way. The fix is to build the system—and then run it with enough discipline that leadership trusts it.

See Fractional COO/CDO for how this type of engagement works. Or book a call to talk through your situation.