1. Abstract
Adjusting multiple assumptions simultaneously can obscure the true impact of individual drivers. Testing scenario sensitivity one variable at a time clarifies causal relationships and improves the credibility of scenario analysis.
2. Context
Apply this best practice when building or reviewing scenario analyses in Foresight, particularly when presenting baseline, optimistic, or pessimistic cases to stakeholders.
3. Content
3.1 Why It Matters
Scenario analysis is most valuable when it clearly demonstrates how changes in specific drivers affect outcomes. When multiple inputs are adjusted at once, it becomes difficult to determine which variable is responsible for forecast variance.
Without isolated testing:
- Sensitivity may be misattributed
- Scenario results may appear inconsistent
- Executive discussions may focus on outcomes rather than drivers
- Risk exposure may be misunderstood
Testing one driver at a time allows you to quantify the individual contribution of each variable before layering assumptions together.
3.2 How to Apply
When designing scenarios:
- Begin with your baseline scenario.
- Select a single key driver (for example, inflation, interest rates, or wage growth).
- Adjust only that driver using Mass Apply Values or manual edits.
- Review and record the change in key outputs.
- Return the driver to baseline.
- Repeat the process for other major drivers.
- Once individual sensitivities are understood, combine drivers into structured optimistic/pessimistic or other custom cases.
3.3 Example
A retailer tests the impact of a 1% increase in interest rates. After adjusting only the interest rate driver, forecast demand declines by 2.5%. This isolated sensitivity provides clarity before adding inflation or wage assumptions into broader downside scenarios.
3.4 Common Pitfalls
- Changing multiple macro drivers at once
- Failing to reset drivers to baseline before testing the next one
- Interpreting combined variance without understanding individual impact
- Not documenting driver-level sensitivity results
3.5 Expected Results
- Clear attribution of forecast variance to individual drivers
- Stronger scenario narratives
- More structured executive conversations
- Improved confidence in risk and opportunity assessment
