Improving Forecast Accuracy Using Simulations
Simulations allow you to perform automated what-if analysis on your Strategic Modeling models, optimizing your decision making.
Simulation allows you to assign a range of possible values to inputs that are uncertain and analyze how that uncertainty affects related accounts. You can:
- Quickly calculate and review hundreds of possible outcomes and the likelihood they will occur
- Analyze best and worst case scenarios
- Assess the likelihood of meeting target goals
- See how key uncertain inputs affect your bottom line
For example, you can use simulations to evaluate an expansion scenario for a Sales division. As part of forecasting overall sales in the upcoming year and identifying the likelihood of reaching targets, you can investigate the profitability of the Expansion scenario in which more retail stores are added. Without a simulation, you’re limited to performing a simple and error-prone what if analysis by entering best guess estimates one cell at a time for uncertain input accounts such as Unit Volume, Services Revenue, and Maintenance Revenue and evaluating their impact on Total Sales Revenue. By running a simulation using a range of assumptions for these input accounts, the model is calculated hundreds of times using random scenarios created from the input assumptions. The simulation results show the range of forecast sales and their likelihood.
Strategic Modeling uses Monte Carlo simulation to randomly generate a range of values for assumptions that you define. For more information about Monte Carlo simulation, see About Monte Carlo Simulation and Simulation Accuracy.
Overview of steps to define and run a simulation:
- From the Home page, click Strategic Modeling, then Models, and then open and check out the model you want to analyze.
-
In the Account View, from the Actions menu, click Simulation.
Users with Modeler permissions can run simulations.
- Determine your key input cells and define them as assumptions. See Defining Assumptions.
- Select output cells that are the target of your analysis and define them as forecasts. See Defining Forecasts.
- Optionally, adjust the simulation settings. See Adjusting Simulation Settings.
- Click Run
to run the simulation.
Note:
-
The model is calculated hundreds of times using random scenarios created from the input assumptions.
-
For each trial of the simulation, the values within assumption and forecast cells are recalculated.
-
Assumptions and forecasts in the model are simulated, not just those in the current account view.
-
Only one simulation can run on a server at a time.
-
- Review the results.
The simulation chart shows the range and likelihood of possible results.
Simulation results are available as long as the model is open and the session hasn't expired.
Simulations are associated with a model, not with a specific scenario.
- To review other metrics, click Add Metric
under the information panels, click the metric to add, and then enter the required parameters.
An information box is added, along with the certainty of achieving the metric.
For example, to add a target value, click Target Value and then enter a label, such as Stretch Goal, and a target value. The information box is updated to show the certainty of meeting the target value. The forecast chart is updated to include a vertical line representing the target value, and is shaded to emphasize the values that meet or exceed the target value.
To adjust parameters for a metric, change the label for an information box, or change the order of the information box display, click Metric Item Menu
in an information box.
You can also work at a later time with a simulation you previously defined. Note that forecasts and assumptions are saved, but simulation results are not saved when you save the model or check it in. See Working with Simulation Objects.
Video
| Your Goal | Watch This Video |
|---|---|
|
Learn about using simulations in Strategic Modeling. |
|
