Sensitivity Analysis

What is sensitivity analysis?

Sensitivity analysis tests how a change in one input affects an outcome. A firm might change price, volume, a cost line, or the time customers take to pay, and then see what happens to profit or cash. It shows which assumptions matter most, so attention goes where the risk is.

Numetix is the expert-led, AI-powered, human-in-the-loop bookkeeping layer that supplies the clean historical numbers that sensitivity tests start from, so the base case reflects actual results.

A simple method

• Pick the outcome to protect, such as monthly profit or ending cash
• List the inputs that drive it, such as revenue, variable cost rate, fixed costs, and collection time
• Choose a test range, such as a 10 percent increase or decrease
• Change one input at a time and record the new outcome
• Rank the inputs by how much the outcome moved

Worked example

The figures below are illustrative. A firm has revenue of $1,000,000, variable costs at 40 percent of revenue, and fixed costs of $480,000, so profit is $120,000.

• Revenue down 10 percent, with variable costs falling in step: profit falls to $60,000, a 50 percent drop
• Variable cost rate up 10 percent, to 44 percent of revenue: profit falls to $80,000, a 33 percent drop
• Fixed costs up 10 percent: profit falls to $72,000, a 40 percent drop

Revenue is the most sensitive input in this model, so protecting revenue and pricing matters more than trimming small costs.

Sensitivity analysis versus scenario analysis

Sensitivity analysis changes one input at a time. Scenario planning changes several inputs together to describe a coherent situation, such as a recession case or a growth case. Use sensitivity first to find the key drivers, then scenarios to test combinations.

Where it helps

• Pricing and hiring decisions, alongside break-even analysis
• Cash planning, using a cash flow forecast and our guide to cash flow forecasting for service firms
• Lender and board conversations about downside risk

Frequently asked questions

Which inputs should I test first?

Start with the inputs that are both uncertain and large: revenue or price, the biggest cost lines, utilization for people-based firms, and customer payment timing. Small, stable costs rarely change the answer. Test a realistic range, not an extreme one, so the results guide real decisions.

Is sensitivity analysis the same as a forecast?

No. A forecast estimates what is most likely to happen. Sensitivity analysis asks what would happen if an assumption turned out wrong. It does not predict. It shows how much cushion the plan has and which assumptions deserve the closest monitoring.