Forecast Analytics

Measure forecast accuracy, compare scenarios, and extrapolate trends to improve the reliability of your value predictions. Forecast Analytics help you answer: "How good are our predictions, and what happens if conditions change?"

Who this is for

End User Admin Executive

Prerequisites

• Approved business case with baseline forecasts.
• At least one period of actuals entered in the Realization Plan.
• Reviewed Core Concepts: Forecasts.

Step-by-step instructions

1. Open Forecast Analytics

1. Navigate to Analytics → Forecasts.
2. Select an account and initiative from the dropdown.

2. Review accuracy metrics

The dashboard shows three primary metrics:

Metric	Description	Healthy range
MAPE	Mean Absolute Percentage Error — average forecast error	Below 20%
Bias	Tendency to over-forecast or under-forecast	Near 0%
Tracking Signal	Cumulative error trend	Between -4 and +4

3. Compare scenarios

1. Select two saved scenarios from the dropdown.
2. The chart overlays both forecast curves on the same timeline.
3. The table below shows period-by-period differences.
4. Use this to answer questions like: "What is the revenue gap between our base case and pessimistic case in Q4?"

4. Extrapolate trends

1. Choose a trend method:

Method	Best for	Requirements
Linear	Stable, predictable growth	3+ data points
Moving Average	Noisy data with no strong seasonality	3+ data points
Seasonal Adjustment	Recurring quarterly or annual cycles	2+ full cycles

1. Set the extrapolation window: 1 to 12 future periods.
2. Review the projected curve and confidence bands.

5. Adjust the forecast

1. If actuals diverge, return to the Value Model tab.
2. Update formula variables or assumptions.
3. Recalculate in the ROI Calculator.
4. Save the updated scenario and document the reason.

6. Export the analysis

1. Click Export Chart to download a PNG.
2. Click Export Data to download a CSV of forecasts, actuals, and errors.
3. Use these in stakeholder presentations or quarterly reviews.

Tip: Use seasonal adjustment for recurring benefits

If your value drivers include quarterly or annual cycles, seasonal adjustment produces more accurate extrapolations. For example, retail cost savings may spike during holiday seasons.

Permissions required

Role	Permission	Scope
User	View / Create scenarios	Assigned initiatives
Admin	Edit tolerances / Configure methods / Set default horizons	Tenant-wide
Executive	View / Export	Portfolio

Limits and guardrails

Limit Forecast analytics require at least 3 data points per driver. Limit Extrapolation is limited to 12 future periods. Limit Scenario comparisons are limited to 2 scenarios at a time.

Troubleshooting

Issue: MAPE is extremely high

Cause: Baselines were set incorrectly, or a one-time event distorted actuals. Resolution: Review the Realization Plan for data entry errors, or exclude outlier periods from the analysis. Check for duplicate actuals.

Issue: Scenario comparison shows identical curves

Cause: The scenarios use the same variable set, or one scenario was not saved correctly. Resolution: Re-create the scenarios with distinct variable values in the ROI Calculator. Ensure you click Save Scenario after each change.

Issue: Seasonal adjustment is grayed out

Cause: Insufficient data cycles exist for the selected driver. Resolution: Enter actuals for at least 2 full cycles (for example, 8 quarters for quarterly seasonality) before using seasonal adjustment.

Issue: Tracking signal is outside ±4

Cause: The forecast model is systematically biased and needs recalibration. Resolution: Update the forecast assumptions in the Value Model tab and review the bias direction. Positive bias means you are under-forecasting; negative bias means over-forecasting.

Related pages

• ROI Analytics
• Trend Analysis
• Value Realization Analytics
• Core Concepts: Forecasts

Escalation path

For forecast calculation errors or missing scenario data, open a support ticket with severity P3.