Generate insights for your multivariate forecasts

Combine predictive analytics and generate AI to better understand your forecasts. When you create a multivariate forecast and have the Planning Analytics AI assistant enabled, you receive AI-generated insights that explain model accuracy and highlight key factors that influence forecast values.

When you generate a multivariate forecast, the summary displays in the Forecast summary tab. 

Previously, you could generate summaries only for univariate forecasts. With this update, you can now generate summaries for multivariate forecasts as well.

Choose a model family for univariate forecasts

Gain more control over your forecasts by selecting a family of Holt-Winters models for univariate forecasts. 

By default, Planning Analytics automatically selects the forecast model that has the high predictive accuracy. If you prefer a particular model output rather than the highest predictive accuracy, you can choose to use a different model from a family of Holt-Winters models, which include the following selections:

• Auto (default)

• Trend

• Additive trend

• Damped trend

• Seasonal

• Additive seasonal

• Multiplicative seasonal

To select a family of models from your univariate forecast, click the Advanced tab, then from Select a model, click the family of models that you want to use.

Perform variance analysis using AI-generated insights

You can now use variance analysis to quickly compare planned financial outcomes with actual financial results to identify the differences and to determine the reasons for these differences.

You can analyze differences between two drivers (such as Unit Price and Unit Sales) across up to 120 rows in a cube view. Launch the analysis by selecting the two columns that represent the two versions (for example, 2024 Actual vs 2025 Actual), then right-click and select Variance Analysis (AI). Choose the relevant drivers, then Planning Analytics displays an AI-generated summary and highlights the top and bottom variances for each driver. Drill down into individual members to explore detailed variance tables and understand which driver had the greatest impact on each product.

Undo and redo actions in cube views

Each cube view maintains its own record of up to 20 steps, so you can roll back up to 20 operations or the same type that you can undo in the dashboard. The steps that are in the record include any data entry changes.

To undo or redo operations, click Undo or Redo on the cube view toolbar.

Update schedules for multiple chores in Database explorer

Save time by adjusting schedules for multiple chores at the same time.

When you select several chores in the Database explorer, click Update chore schedule in the toolbar to open a scheduling dialog. You can apply a single schedule to all selected chores or shift each chore's time relative to its current schedule (for example, to adjust your chore schedules for daylight saving). This streamlined batch scheduling approach makes managing chore execution more efficient.