What it means
Every forecast rests on assumptions: how many customers will sign up, what they will pay, what raw materials will cost, how long a project will take. Sensitivity analysis tests these one by one, holding everything else steady, and records how far the profit, cash balance or valuation shifts in response.
The output is usually expressed as an elasticity or a simple table. If cutting the price assumption by 5% wipes out half the profit, price is a high-sensitivity variable; if a 5% change in office rent barely registers, rent is a low-sensitivity variable that does not deserve another week of research.
In practice this reshapes decisions rather than just decorating a spreadsheet. A team that knows profit is highly sensitive to customer churn will invest in retention rather than in shaving supplier costs, because the model has told them where the leverage sits.
Sensitivity analysis is often confused with scenario analysis, but they are different tools. Sensitivity analysis moves one variable at a time to isolate its effect, while scenario analysis moves a coherent bundle of variables together to describe a plausible future such as a recession or a competitor price war.
The main limitation is that real variables rarely move alone. A recession that reduces sales volume usually also softens input prices and wage growth, so a one-at-a-time test can exaggerate the pain or the relief of a single change.
In practice
Real-world examples.
Example
A subscription analytics firm models its cash runway and finds that a 2 point rise in monthly churn shortens runway by five months, while a 10% rise in salaries shortens it by one month. The board redirects the quarterly plan towards onboarding and customer success rather than hiring freezes.
Example
A construction contractor bidding for a $6,000,000 fit-out tests steel price assumptions across a range of plus or minus 15%. The analysis shows the job turns loss-making above a certain price, so the contractor writes a materials escalation clause into the bid.
Example
A hotel group preparing a refurbishment case flexes occupancy, average room rate and build cost separately. Average room rate proves the most sensitive input, so the team commissions extra market research on pricing before committing to the spend.
Think of it
“Sensitivity analysis is like testing how a recipe changes when you vary ingredients. You discover that doubling sugar changes the taste dramatically.
Formula
Calculation
Sensitivity = % Change in Output / % Change in Input
Start with a base case for a small equipment hire business. Revenue is $2,000,000, variable costs run at 60% of revenue, and fixed costs are $500,000.
Base case operating profit: $2,000,000 - $1,200,000 - $500,000 = $300,000.
Now flex revenue down by 10%, holding the cost structure constant. Revenue becomes $1,800,000 and variable costs fall to $1,080,000, because they scale with sales.
New operating profit: $1,800,000 - $1,080,000 - $500,000 = $220,000.
The profit fell by $80,000, which is $80,000 / $300,000 = 26.7% of the base. Dividing by the 10% revenue change gives a sensitivity of about 2.7, meaning every 1% move in revenue swings operating profit by roughly 2.7%. That figure tells the owner that defending revenue is worth far more than trimming small fixed costs.Case study
Seen in the real world.
This is an illustrative, fictional case. Northwind Ceramics was a mid-sized tableware maker preparing a business case for a new $1,400,000 kiln. The finance lead's model showed a healthy return, and the operations director was ready to sign.
Before approving it, the managing director asked for a sensitivity table across four inputs: unit selling price, energy cost, output volume and the useful life of the kiln. Three of the four barely moved the answer, but energy cost was punishing, with a 20% rise turning a comfortable return into a marginal one.
Northwind still bought the kiln, but the sensitivity work changed the terms of the deal. They negotiated a three-year fixed-price energy contract before installation and added a small solar array to the roof of the works. The investment went ahead with its single biggest risk contained rather than merely noted.
Watch out
Common mistakes.
- Flexing several assumptions at once and calling it sensitivity analysis, which makes it impossible to see which input drove the change in the result.
- Testing tiny, comfortable ranges such as plus or minus 1%, so the analysis never reveals the point at which the decision would actually flip.
- Treating the model output as a forecast rather than as a map of risk, and then presenting the worst sensitivity case to the board as if it were expected.
Questions
People also ask.
How wide should the test range be?
Wide enough to cover what has genuinely happened before in your market, which for most cost inputs means at least plus or minus 10% to 20%.
Is sensitivity analysis the same as stress testing?
Not quite, because stress testing pushes assumptions to deliberately severe levels to check survival, while sensitivity analysis measures responsiveness across a normal range.
Which variables should I always test?
Price, volume and any single cost that makes up more than about 10% of your total costs, since these almost always dominate the result.
From the founder's library
Take it further with the book.
Build your financial confidence beyond this definition. Shihan's full-length guide, Accounting Fundamentals, takes the same plain-English approach and turns it into a complete, practical playbook for non-finance managers, business owners and students - with chapter-end quiz answers and presentation slides included.
25% off with code MMHQ25, applied at checkout. Priced in USD - checkout may show the equivalent in your local currency.
View the book and save 25%