How to budget, re-forecast and plan scenarios without wasting time.

Once you have a basic financial model, the hard part is using it. The how-to on P&L, balance sheet and cash flow forecasts explains how to build the three statements. This guide covers how to run the business with them: setting an annual budget, keeping a rolling forecast, comparing both with actuals and planning for scenarios.

Budget versus forecast

• The budget is set once a year. It records what you have agreed to spend and what revenue you are aiming for. It is a yardstick, not a prediction.
• The rolling forecast is updated every month or quarter and always looks 12 to 18 months ahead. It reflects what you now expect, given what you have learned.

Comparing actuals with the budget shows whether you are on plan. Comparing the forecast with the budget shows whether the plan is still realistic.

Build the budget from drivers

1. Start with revenue drivers. For a SaaS business: new customers per month, average revenue per customer, churn and expansion. For services: billable days, day rates and utilisation. Tie new customers to pipeline and conversion assumptions from your sales function.
2. Build people costs line by line. Each role, start month, salary, employer National Insurance and pension. People are usually the largest cost, so this is where accuracy matters most. The hiring and team cost planner helps here.
3. Add cost of sales: hosting, third-party services, payment fees and any delivery effort per customer.
4. Add overheads: software, premises, insurance, professional fees, marketing.
5. Link everything to cash timing so the budget produces a monthly cash balance, not only a P&L.
6. Agree spending authority. Who can approve spend, up to what amount, and when a variance needs a conversation.

Run the monthly cycle

After each monthly close, put actuals next to budget and forecast. Look only at lines that differ materially; a rule of thumb is more than ten percent or more than a fixed amount that matters to you. For each, ask whether it is timing (it will catch up) or a real change (update the forecast). Then roll the forecast forward a month.

Plan three scenarios

Keep a base case, a downside and an upside. The downside should be genuinely uncomfortable: slower sales, a big customer leaving, a delayed funding round. For each, write down in advance the actions you would take and the trigger that would make you take them. This turns a crisis into a prepared response. A short list of pre-agreed cuts is worth more than a perfect model.

Worked example

Ferncast is a fictional Leeds tech startup selling compliance training software to care homes. Its annual budget assumed eight new customers a month and three hires during the year. After four months the rolling forecast showed new customers averaging five a month because care home decisions were slower than expected, while churn was lower than planned.

Rather than keep spending to the original budget, the team updated the forecast, held the third hire until new customers reached seven a month for two consecutive months, and moved part of the marketing budget into a referral scheme with existing customers. The downside scenario, prepared at the start of the year, already listed the hire freeze as the first lever, so the decision took one meeting rather than weeks of debate.

Common mistakes

• Building a top-down revenue line (“we will grow ten percent a month”) with no link to pipeline or conversion.
• Spending to budget when the forecast says revenue will be lower.
• Forgetting employer National Insurance, pension contributions, recruitment costs and equipment for new hires.
• A model so complex only one person can update it.
• Only one scenario, which is usually the optimistic one.

What to measure

• Revenue and cost variance to budget and to last forecast.
• Forecast accuracy: how close last quarter’s forecast was to what happened.
• Cash at the low point in each scenario.
• Progress against the trigger points you set for hires and cuts.