Introduction to Cases
How to use Scenarios and Sensitivities to test your Model
Last updated
How to use Scenarios and Sensitivities to test your Model
Last updated
We often want to test how our Model outputs change for different values of inputs. What would be the impact of raising the price of a product? What would happen if our product launch was delayed by 6 months?
To do this, we can use Scenarios and Sensitivities:
Models automatically calculates all combinations of Scenarios and Sensitivities and, unlike Excel, can easily display several Scenarios or Sensitivities simultaneously, as in the example screenshot below, enabling efficient analysis of your Model dynamics.
We call each combination of a Scenario and Sensitivity a Case.
Scenarios and Sensitivities both affect the values of Assumptions in your Model. For each Scenario/Sensitivity in the Model, you can apply a value for each Assumption.
In the below example, there are two Scenarios (Base
and Lean
), and three Sensitivities (Base
, Low
, and High
).
In this case, the Lean
Scenario is configured to take its value from the Base
Scenario by default, unless a value is specified for Lean
(1). This means that a value must be present in the Base
Scenario for all Assumptions.
Sensitivities do not need to have a value specified, in which case the value for the Assumption will be the unchanged Scenario value. In the case that a value is specified (2), the Scenario value will be modified appropriately.
The next sections explain how to set up Scenarios and Sensitivities, and the various ways in which they can be used.
Variable previews and Dashboard items will by default display values for the currently active Scenario and Sensitivity. You can change this in the Model Settings bar at the top of the Model Editor. Just click on one of the Scenario or Sensitivity buttons, as shown in the screenshot below, to activate it.
Scenario
A Scenario usually reflects some kind of "alternative reality" for your Model, perhaps to reflect decisions that need to be made, and may involve significant changes.
e.g. turning a revenue item on or off, or switching between insourcing and outsourcing a cost activity
Sensitivity
A Sensitivity is usually used to test how the outputs of a Model are affected by changes to the inputs, reflecting uncertainty in our assumptions. e.g. testing the impact of a range of inflation rates