Forecast Quality KPI Insight

Use the Forecast Quality KPI Dashboard to view a consolidated overview of forecast quality across a dataset. The dashboard presents key performance indicators, method distribution, and competitive model health in a single view

Demand planners and administrators previously had no centralized way to assess overall forecast health after a calculation run. Evaluating forecast accuracy, method diversity, bias, and competitive model performance required reviewing individual series or running separate queries. Demand planners and administrators needed a summary-level view to quickly identify areas that required attention.

These are the components of the DMP Forecast Quality KPI Dashboard:

• Six KPI summary cards showing Total Series, Average sMAPE, Poor Fit count, Bias Alerts, Method Diversity, and Competitive coverage percentage, with color-coded thresholds (green, orange, and red) indicating health status.
• A Health Score gauge displays a score from 0 to 100 based on the overall sMAPE and classifies forecast performance as Excellent, Good, Fair, Poor, or Critical.
• A Wins per Method bar chart showing the number of series for which each forecast method is the best fit, highlighting method dominance or diversity issues.
• An Assignment Breakdown table showing series distribution across No Method, Explicit Method, Competitive Forecast, and Ensemble Winner categories.
• A Quality per Competitive Model table displaying average sMAPE, poor fit count, and bias count per competitive model with traffic-light indicators.
• A Winner Method Distribution table showing each winning method's win count, average sMAPE, and quality rating.
• A drill-through path to Forecast Series Explorer on Assignments, Poor Fit / Bias, and Method Wins.