What-If Analysis

What-if analysis is a way of exploring a decision by changing the assumptions behind a model and watching how the result responds. You build a model — a budget, a forecast, a pricing plan, a return projection — that turns inputs into an outcome, then ask a series of what-if questions: what if sales grow slower, what if costs rise, what if we hire two months later, what if the conversion rate is half our estimate? Each change ripples through the model to a new result, and comparing those results shows how the outcome depends on the assumptions. Also called scenario analysis, it moves planning from a single guessed number to a range of possibilities, revealing which assumptions the outcome is sensitive to and which barely matter. It is less about predicting the future than about understanding how a plan behaves across the futures that might happen.

The value of what-if analysis is that it exposes risk and dependence before money is committed. A plan that looks fine on one set of assumptions might collapse if a single input moves the wrong way, or it might prove robust across a wide range — and you only learn which by varying the inputs and looking. This is why it underpins budgeting, financial modeling, capacity planning, and investment decisions: it turns a fragile point estimate into a tested range, highlights the assumptions worth pinning down, and surfaces the downside scenarios worth preparing for. A best case, a base case, and a worst case together tell a far richer story than a single forecast, because they show not just where the plan is aimed but how much room there is if reality disagrees.

What-if analysis is closely related to sensitivity analysis, and the two are often used together. Sensitivity analysis is the narrower, more systematic cousin: it varies one input at a time to measure how much the outcome moves, isolating which single assumptions the result is most sensitive to. What-if analysis is broader — it can change several inputs at once to build whole scenarios (a recession case, an aggressive-growth case), not just measure the impact of one lever. In practice you use sensitivity analysis to find which assumptions matter most, then what-if scenarios to combine those assumptions into coherent stories about how things might unfold. One asks how much each dial affects the outcome; the other asks what the world looks like when several dials move together.

What-if analysis also connects to pro forma financials, but they play different roles. Pro forma financials are the modeled statements themselves — the projected or adjusted numbers built on a set of assumptions. What-if analysis is the act of flexing those assumptions to generate alternative versions of them. You might build a pro forma projection on a base case, then run what-if analysis to see the pro forma results under slower growth or higher costs, producing a family of pro forma statements rather than one. So the pro forma is the output, and the what-if is the exploration that stress-tests it. Keeping the roles clear matters: a single pro forma projection presented without any what-if testing hides how fragile or robust its assumptions really are.

Using what-if analysis well means varying the assumptions that actually matter and being honest about the range. Start by identifying the key drivers — the handful of inputs the outcome truly depends on, often found with sensitivity analysis — and build scenarios around them rather than fiddling with trivial ones. Construct a genuine spread: a plausible best case, a realistic base case, and a worst case severe enough to be uncomfortable, so the downside is actually tested rather than gestured at. Use the results to make decisions more robust — to size a cash cushion, choose a plan that survives the bad scenario, or pin down the assumptions worth researching further. The goal is not a single confident forecast but a clear-eyed view of how the plan behaves across the futures it might face.

The failures come from doing it carelessly or not at all. Committing to a plan on a single point estimate, with no what-if testing, hides how sensitive it is and how badly it can break. Running only optimistic scenarios — a best case and a slightly-less-best case — tests nothing and breeds false confidence. Varying inputs that barely move the outcome while ignoring the ones that dominate wastes effort on the wrong dials. And treating the scenarios as predictions rather than as a range of possibilities invites surprise when reality lands outside the tidy cases modeled. The discipline is to flex the assumptions that matter, build an honest spread including a genuinely bad case, and use the results to choose plans that hold up across scenarios rather than plans that only work if everything goes right.

What-if analysis — varying a model's inputs to see how the outcome moves — turns a single forecast into a tested range of scenarios, exposing which assumptions a plan depends on and how it behaves under stress.

Etymology: source.

Search interest for this term over the last five years:

View interest-over-time on Google Trends →

Areas of marketing where what-if analysis is a core concern: