Verified against ChatGPT · 2026-08-14
Build a what-if scenario model that shows its assumptions instead of one confident-looking number
Structures a scenario/what-if analysis around explicit, adjustable assumptions and a spread of outcomes, so the output is a decision tool rather than a single fabricated-precision forecast.
The prompt
Ready to copy — highlighted parts are example details you can swap.
Build a scenario analysis for the decision below, structured around explicit assumptions rather than a single point estimate that hides how sensitive the answer is to what you assumed. DECISION BEING EVALUATED Whether to open a second fulfillment warehouse in the Midwest next year KEY VARIABLES THAT DRIVE THE OUTCOME Order volume growth rate, average shipping cost saved per order, warehouse setup and lease cost, ramp-up time to full efficiency BASELINE DATA OR CURRENT STATE Current order volume growing 12% YoY, current average shipping cost per order is $6.40, a comparable warehouse elsewhere took 5 months to reach efficiency TIME HORIZON 24 months from opening STRUCTURE Identify the two or three variables in Order volume growth rate, average shipping cost saved per order, warehouse setup and lease cost, ramp-up time to full efficiency that the outcome is actually most sensitive to — not every variable deserves its own scenario axis, so name which ones matter enough to vary and which can be held at a reasonable fixed assumption instead. State every assumption as an explicit, numbered input (growth rate, cost per unit, adoption rate — whatever applies) with the specific value used, not folded invisibly into a formula, so anyone reviewing this can see exactly what was assumed and challenge any single input without redoing the whole model. Build three named scenarios — conservative, base case, and optimistic — each defined by a specific, stated combination of the key variables at low/expected/high values, not just three outcome numbers with no visible inputs behind them. Show the actual outcome number for each scenario, and state which scenario the current trajectory most resembles based on Current order volume growing 12% YoY, current average shipping cost per order is $6.40, a comparable warehouse elsewhere took 5 months to reach efficiency, so the analysis connects to where things stand today rather than floating as three abstract hypotheticals. Identify the single assumption the outcome is most sensitive to — the one where a modest change in that one input moves the final number more than an equivalent change in any other input — since that's the one variable worth watching or de-risking most closely. WHAT NOT TO DO Do not produce a single "expected value" number as the headline without the three scenarios and their assumptions visible alongside it — a lone number invites false confidence in what is actually a range. Do not vary every input across every scenario if the sensitivity check shows most of them barely move the outcome; that dilutes attention away from what actually matters. OUTPUT FORMAT 1. The 2-3 variables chosen as scenario axes, and why the others were held fixed. 2. Table: scenario | key variable values | resulting outcome. 3. Which scenario current data most resembles. 4. The single most sensitive assumption, and what a plausible swing in it does to the outcome. 5. One paragraph translating the range into a decision recommendation.
Customize
Optional — swap in your own details for the highlighted parts above.
Why this works
A model asked to "model this decision" will often produce a single confident number, because a point estimate reads as more decisive and complete than a range with visible assumptions, even though the single number is actually less honest about what's genuinely known versus assumed — requiring three named scenarios with explicit, numbered assumption values forces the model to expose exactly what it's assuming rather than let those assumptions disappear into an opaque calculation, which is the entire point of a scenario analysis as a decision tool rather than a forecast. Requiring the sensitivity identification — which single assumption moves the outcome most — targets a specific and valuable piece of information that a flat three-scenario table doesn't automatically surface: knowing that the warehouse decision is mostly sensitive to ramp-up time, and much less sensitive to the shipping-cost assumption, tells the decision-maker exactly which number is worth spending more effort de-risking or verifying before committing, versus which ones can be left as reasonable estimates. Anchoring the scenarios explicitly to which one the current baseline data most resembles prevents the common failure of a scenario analysis floating as three disconnected hypotheticals with no bridge back to where things actually stand today — without that anchor, a reader has no way to judge whether the "optimistic" scenario is a stretch or already roughly on track, which is precisely the judgment the analysis exists to support. Limiting scenario axes to the two or three variables that actually matter, rather than varying every input, reflects a deliberate anti-complexity choice: a model given free rein will often vary every listed variable simultaneously to seem thorough, producing a combinatorial mess of scenarios that obscures rather than clarifies which few things the outcome is actually riding on.
What you get back
Scenario axes chosen: order volume growth rate and ramp-up time to efficiency — setup/lease cost held fixed since it varies little in practice. Conservative: 8% growth, 8-month ramp → 24-month net savings: $410K. Base case: 12% growth, 5-month ramp → 24-month net savings: $780K. Optimistic: 16% growth, 3-month ramp → 24-month net savings: $1.15M. Current trajectory most resembles: base case, given the 12% YoY growth already observed. Most sensitive assumption: ramp-up time — a 2-month delay beyond the base case erases roughly 35% of projected savings, more than an equivalent swing in growth rate would.
Verified against
ChatGPT GPT-5.1 · 2026-08-14
Changelog
- 2026-08-14 — Initial publish, verified against ChatGPT GPT-5.1.
Need this built into your business?
If a prompt isn't enough — what Scult builds, built and maintained for you — that's Scult's day job.
EXPLORE WHAT SCULT BUILDS
