The Forecast That Never Gets Revised
A SaaS company's FP&A team produces a 13-week cash forecast. For six consecutive weeks, actual cash collections have come in 8–12% below forecast, always in the same direction. The model assumes a 45-day DSO based on historical averages, but the sales mix has shifted toward enterprise customers with 60–75 day payment terms. The team's process is: sales enters expected close dates, finance applies a fixed 90% probability to all pipeline, and no one revisits prior forecasts. The controller wants to know why the forecast is 'always wrong' and whether the 13-week horizon is even the right tool, given the company has 9 months of runway.
1. Diagnose the systematic forecast error and identify the specific modeling flaws.
Read the worked answer
The error is biased, not random - always the same direction - which points to structural flaws:
- Stale DSO assumption: A blended 45-day DSO no longer matches the mix; enterprise terms of 60–75 days mean collections systematically arrive later. The average is being dragged by a shrinking SMB base.
- Flat 90% probability: Applying one probability to all pipeline ignores that early-stage deals close far less often than late-stage. This overstates near-term collections.
- No feedback loop: Forecasts are never compared to actuals, so the bias is never corrected. A forecast without a variance review is a guess with formatting.
Fix: segment DSO by customer tier, apply stage-weighted probabilities (e.g., 30/50/80%), and institute a weekly forecast-vs-actual variance review with a documented adjustment factor.
2. Given 9 months of runway, is a 13-week cash forecast the right instrument? What would you recommend instead or in addition?
Read the worked answer
A 13-week direct cash forecast is the right operational tool for near-term liquidity management, but it is insufficient as the only tool with 9 months of runway, because the risk horizon is longer than the forecast horizon. Recommend a layered approach:
- 13-week direct forecast for weekly cash management and covenant monitoring - keep it, but fix the bias.
- Rolling 12-month indirect (cash flow statement) forecast to capture the runway question, including seasonality and the enterprise-collections lag.
- Scenario/stress analysis: base, downside (collections slip another 15 days), and a covenant-breach case. With 9 months of runway, the downside scenario likely shortens it materially - quantify that.
- Trigger points: define early-warning thresholds (e.g., DSO > 65 days, runway < 6 months) that force action.
The 13-week tool answers 'can we make payroll this quarter'; the 12-month scenario answers 'do we need to raise or cut, and when.'