Configure outlier dates to improve accuracy

Outlier dates, or “non-representative days,” are used to exclude specific dates from historical sales analysis to ensure more accurate budgeting and forecasting. These typically include holidays, local events, or operational anomalies that cause unusual sales patterns.

Start from: The Analytics Dashboard.

• Navigate to Configuration > Outlier Dates.

• Select the Year and the Site for which you want to configure non-representative dates. These dates are set on the current year's calendar but are used to adjust the sales budget for the following year.

Example: If you mark dates in 2024 as outliers, they will not influence the sales budget calculations when 2024 is used as the historical reference for 2025.

Review system-suggested outliers

• Days outlined in red are Potentially Non-Representative Days. These are flagged automatically based on the sensitivity setting and variations detected in the data source (POS or cash sheet).

• Adjust the sensitivity using the slider control to detect fewer or more outliers.

Manually mark non-representative days

• Click on any date to mark it as Non-Representative.
• The selected day will appear in yellow.
• Add a comment to justify the selection (e.g., “Bank holiday” or “Local event”).

• Select Accept to confirm or Cancel to discard.

Generate a report

To get a global overview of all configured outlier dates for a specific site:

• Select Report at the top of the grid.
• A PDF summary will be generated listing all the dates marked as Non-Representative for the selected site and year. This report helps provide visibility and traceability for forecasting decisions across your organisation.

Originally published at https://help.mapal-os.com/analytics/configuration/outlier-dates/configure-outlier-dates-to-improve-accuracy/