Every decision about hiring, pricing, marketing, or investment starts with the same question: what do the numbers look like under this assumption? Financial forecasting is the discipline of answering that question well enough to actually use - not perfectly, but reliably enough to make better decisions.
Financial forecasting - the practice of projecting your business's future revenue, expenses, profit, and cash position based on current data, historical patterns, and explicit assumptions about what will change.
Three horizons, three purposes
Different decisions need different forecasts. A small business typically maintains three.
Short-term (1-3 months)
Weekly resolution. Operational planning. The cash flow forecast lives here. Used for "can we cover payroll next Friday?" and "should we accept these payment terms?" Accuracy target: ±5-10%.
Medium-term (3-12 months)
Monthly resolution. Tactical planning. Used for "can we afford this hire?", "will we hit the year-end target?", and "should we invest in this campaign?" Accuracy target: ±15-20%.
Long-term (1-3 years)
Quarterly resolution. Strategic planning. Used for major capital decisions, hiring roadmaps, and investor or lender conversations. Accuracy degrades fast at this horizon - treat it as directional rather than precise.
Forecast vs budget
Two related but distinct concepts often confused:
- Budget - what you commit to spending. A plan. Anchored at the start of the period and updated sparingly. Used to enforce discipline.
- Forecast - what you expect will happen. A prediction. Updated continuously as new information arrives. Used to anticipate.
Both have a place. A business that operates without a budget drifts; a business that updates only its budget and not its forecast operates blind to changes mid-year.
Common forecasting methods
Top-down
Start with a market or growth-rate assumption, work down. "The market is $X billion, we want 0.5% share by year 3, that's $Y of revenue." Useful for new ventures and strategic planning; usually too aggressive for operational use.
Bottom-up
Start with current activity, build up by category. "We had 200 customers last month, 5% monthly churn, 15 new customers expected, at $X average revenue." More accurate for short and medium horizons because it grounds in actual data.
Driver-based
Identify the 2-3 metrics that actually drive revenue (customers × revenue per customer; visitors × conversion × average order), forecast each driver, multiply. The most diagnostic method - when reality differs, you can identify which driver was off.
Trend extrapolation
Take the growth rate from the last 6-12 months, project forward. Simple, often good enough for stable businesses, misleading for businesses with changing dynamics.
What to forecast (always)
A complete forecast covers four things, not just revenue.
- Revenue - by major segment or product line if material
- Expenses - split between fixed (rent, salaries, software) and variable (cost of goods, payment processing, contractor costs)
- Profit - gross, operating, net (see Net Profit Explained)
- Cash flow - the timing of receivables and payables, which can differ significantly from profit
Revenue-only forecasts are the most common mistake. The number that tells you whether the business will work is rarely revenue alone.
Assumptions matter more than the numbers
Every forecast rests on explicit assumptions. The discipline of writing them down is what separates a useful forecast from a number on a screen.
When reality differs from forecast, the diagnosis is usually "which assumption was wrong?" - and that's only answerable if the assumptions were stated upfront.
Common forecasting mistakes
1. Overestimating the near term
Almost every business overestimates how fast the next quarter will move. Things take longer than you think; deals close slower; hires ramp later. Build in slack.
2. Forecasting in straight lines
Real businesses have seasonality, lumpy deals, calendar effects. A forecast that grows 5% every month for 12 months will always be wrong somewhere.
3. Treating the forecast as the goal
A forecast is a prediction, not a target. Confusing the two leads to incentives where people massage the forecast to match performance instead of using it to plan.
4. Never revisiting
A forecast built once and never updated is a fairy tale. Update monthly with actuals.
Related concepts
- Revenue Forecasting Methods - the deeper dive on revenue specifically.
- Expense Forecasting - the expense-side companion.
- Scenario Planning Explained - building multiple forecast versions for different assumptions.
- Cash Flow Forecasting - the short-term cash variant of forecasting.
- Financial Forecasting for Small Businesses - a longer guide to the discipline.