Business Forecasting Best Practices And Methods

Updated 11 October 2026

By Richie Lennon

Business forecasting estimates what your business is likely to earn, spend and need in cash, using explicit assumptions about sales, costs and timing. A budget sets the plan; a forecast tests the likely outcome as conditions change. Start with reliable actual figures, separate sales targets from the baseline, model a downside and update the forecast against results. Its value is the decision it helps you make before committing money.

A proposal in the pipeline is not money available for payroll. Yet it is easy to put an ambitious sales number in the forecast and start spending against it.

We make a clear distinction: the team’s sales target and the owner’s operational baseline have different jobs. Optimism can motivate a team. The spending plan needs assumptions you can explain and a way to cope if the sales arrive late.

We would start with the decision. Can you afford a hire? How much stock should you order? What must happen before you expand? A forecast that cannot help answer the question has become paperwork.

What business forecasting includes

The financial forecast connects expected activity to revenue, costs, profit, assets, liabilities and cash. The level of detail depends on the decision and the business.

Profit and cash are connected but different. You may earn a margin before a customer pays, or pay for stock before it is sold. A loan adds cash but is not sales revenue.

Our cash-flow forecasting guide covers the detailed cash timetable. This guide covers the assumptions and methods behind the wider business forecast.

Forecast, budget and target

A budget records the agreed plan and spending limits. A target describes an outcome the business wants. A forecast describes the outcome currently expected under its assumptions.

They can use the same financial structure without having identical numbers. If a sales opportunity slips, keep the target visible and update the operational forecast. Hiding the change preserves a spreadsheet while making the spending decision worse.

Retain the earlier forecast version so you can compare expectation with outcome. Replacing last month’s prediction without a record makes it difficult to learn why it missed.

Build the baseline from evidence

Start with reconciled recent accounts, the current cash position and known commitments. Explain unusual items before using them as a repeatable run rate.

Separate contracted work, recurring business, likely opportunities and speculative growth. Use your own conversion and timing evidence. A pipeline stage labelled “90%” is not proof the opportunity will close this month.

Include delivery capacity. An agency cannot deliver unlimited projects with the same team. A retailer cannot sell unavailable stock. A software business needs to distinguish new customers, expansion, churn and the price each group pays.

Record who owns each assumption and when it was last reviewed. Sales, operations and finance should challenge one another’s inputs before the owner approves a material commitment.

Choose a method that fits the question

Moving averages

A simple moving average uses recent periods as a starting point. Monthly sales of €45,000, €48,000 and €51,000 average €48,000.

A weighted average gives selected periods more influence. Weights of 20%, 30% and 50% on those months give €48,900: €9,000 plus €14,400 plus €25,500.

Neither figure automatically predicts April. A promotion, a lost customer or seasonality can make the pattern misleading. Explain the reason for the weights rather than choosing them to produce the preferred answer.

Trends and seasonality

Trend analysis extends a pattern in the data. Seasonal analysis considers recurring changes such as buying cycles or holiday demand.

Compare like periods and identify one-off events. Do not extend a three-month growth streak indefinitely or assume last December’s promotion will repeat at the same margin.

Driver-based forecasting

This connects the forecast to the activity producing the result. A retailer might model orders multiplied by average net order value, then product and fulfilment costs. An agency might model deliverable project hours, agreed fees and staffing.

For a recurring-revenue business, model opening customers, new customers, losses and changes in value separately. Multiplying revenue by churn is not a complete sales forecast.

Driver models help explain a change. If sales fall, you can distinguish fewer orders from a lower selling price, rather than applying one unexplained percentage to revenue.

Statistical models and AI

Regression and other statistical methods can explore relationships where there is enough relevant data. AI tools can help inspect data or suggest assumptions, but they do not establish that a relationship will continue.

Check the input, assumptions and calculation. A confident explanation is not evidence that the result is reliable. Begin with a model the team can understand before adding complexity it cannot maintain.

Turn the sales forecast into profit and cash

Here is a simplified one-month illustration. Revenue and profit figures exclude VAT; the cash figures are assumed actual gross amounts due in that month, with tax and other commitments included in outflows.

The month shows €10,000 operating profit but uses €10,000 cash. The difference depends on the actual collection and payment timing and any balance-sheet movements. This is an illustration, not a complete three-way model or client case.

If €10,000 of the expected receipts moves into the following month, closing cash falls to €5,000. Annual profit has not necessarily changed, but the next payroll decision has.

Use separate VAT and tax assumptions in a full model. Avoid mixing sales excluding VAT with receipts including VAT without showing the liability and payment.

Scenarios and sensitivities

A scenario combines a plausible set of conditions. A downside might include slower sales, delayed receipts and a higher supplier cost together. A sensitivity changes one assumption to see how much it matters.

Do both around the decision. For a hire, test the start date, full employment cost, training period, available work and collections. For stock, test sell-through, supplier deposits, lead times and returns.

Set an action alongside the result. If cash falls below the reserve you have agreed, what spending pauses, who follows up receipts and when does the owner decide? A red cell without an action is only a warning.

Choose the horizon around the commitment

Short-term weekly cash forecasts are useful when payment timing is tight. A rolling monthly forecast can cover the next year and keep seasonal commitments visible. Longer-term projections help examine strategy, investment and funding.

Precision usually falls as the horizon extends. That does not make longer planning useless; it means assumptions and scenarios need more attention. Short-term forecasts still require a strategic decision when a commitment has long consequences.

For a funding application, check the specific lender or scheme requirements. LEO Kildare’s projections guidance requests three-year monthly cash and profit projections for its grant application process. That is a specific requirement, not a universal rule for every Irish lender.

Tools should support the process

A well-controlled spreadsheet can be enough to start. Keep assumptions separate, label units and dates, check formulas and control versions.

Dedicated tools can connect accounts and organise scenarios. For example, Fathom’s forecasting documentation describes linked profit, balance-sheet and cash forecasts. Check the required integrations, plan and model before choosing a product.

The tool cannot know an unsigned contract will slip or that a supplier wants a deposit early unless someone changes the assumptions. Connected data helps maintain the model; ownership makes it useful.

Review what changed each month

Compare actuals with the forecast you made, then separate volume, price, margin and timing differences. Update future assumptions without rewriting the historical prediction.

We would keep the review focused:

1. What changed, and why? 2. Does the change repeat or was it one-off? 3. What happens to profit, cash and capacity? 4. Which decision changes, who owns it and by when?

There is no universal “85% accurate” pass mark for every business forecast. Define the measure and the material errors. Getting annual sales close while missing the month’s cash low point can still lead to the wrong commitment.

What Around Finance adds

The useful finance work joins reliable accounts, commercial assumptions and the decision the owner is making. It includes challenging the baseline, explaining the cash consequences and following up what changed.

Our management accounting service is the starting point for regular reporting and decision support. If you run a growing Irish business turning over €250,000 or more, a Finance Fit Call can establish the right scope of ongoing support. Agree any forecasting model or specialist funding work within that scope.

FAQs