Cash flow forecasting is the single most requested — and most frequently botched — financial planning exercise for small and mid-market organizations. The challenge is not conceptual; everyone understands that cash in must exceed cash out. The challenge is structural: which model fits the data you actually have, the decisions you actually need to make, and the time horizon you actually care about.
Three approaches, three use cases
There is no universal cash flow model. The right choice depends on forecast horizon, data availability, and how the forecast will be used. Here are the three approaches we deploy most often with mid-market clients.
Direct method
The direct method forecasts specific cash inflows and outflows: customer collections on a receivables aging schedule, vendor payments from an AP ledger, payroll dates, loan repayments, tax installments. It is granular, accurate over short horizons, and immediately actionable — the CFO can see exactly which week the balance dips and why.
The limitation is effort. Building a direct forecast from scratch requires detailed transaction-level data and assumptions about collection timing that degrade rapidly beyond eight to twelve weeks. For organizations with reliable AR and AP aging data, the direct method is the best short-term tool available.
Indirect method
The indirect method starts with projected net income and adjusts for non-cash items (depreciation, amortization, provisions) and working capital movements (changes in receivables, payables, inventory). It is faster to build, aligns naturally with the P&L budget, and extends comfortably to twelve-month and multi-year horizons.
The tradeoff is precision. The indirect method tells you roughly where cash will land in six months; it cannot tell you whether you’ll miss a payroll run three Fridays from now. For strategic planning, covenant compliance, and board reporting, the indirect method is efficient and adequate. For operational cash management, it is not.
Hybrid method
The hybrid approach layers a direct forecast for the next four to eight weeks on top of an indirect forecast for the remaining horizon. The near-term window captures the granularity needed for day-to-day cash decisions; the outer window provides the strategic view without the maintenance burden of a twelve-month direct model.
The hybrid model is the one we recommend for most mid-market organizations. It matches the level of precision to the level of uncertainty — high detail where you can see clearly, broad strokes where you can’t.
Building with limited historical data
SMEs and younger organizations often have less than two years of reliable financial data, which makes statistical trend analysis fragile. The workaround is structured assumption-building. Interview the sales team about pipeline timing. Review AR aging by customer segment. Map the seasonality of payables against supplier payment terms. These qualitative inputs, structured into a spreadsheet model, produce forecasts that are more accurate than regression on thin data sets — and far more actionable.
The key is to version your assumptions. Label each forecast with the date it was built and the assumptions that drove it. When actuals diverge from forecast, you can trace the variance back to a specific assumption — “we expected Customer A to pay in 30 days; they actually paid in 52” — and refine the model. Over time, the assumptions sharpen, the forecast improves, and the team develops genuine forecasting competence rather than spreadsheet ritual.
Cash flow forecasting is not a one-time exercise. It is a rolling discipline — updated weekly for the direct window, monthly for the indirect window, and reviewed rigorously against actuals every cycle. Organizations that treat it this way rarely find themselves surprised by a liquidity shortfall. The ones that build a forecast once a year and file it in a shared drive learn the hard way.
