Forecast Accuracy Monitoring

With the same underlying concept that has led Analytics research laboratory in building both ibs Analytics and ibs AMS solutions, transparency in accuracy results has been integrated within the solution.

ibs AMS stores and analyzes results from previous forecast simulations in order to retrieve accuracy levels transparently to business users.
For example, considering a daily forecast process embedded in a retail price optimization process.

ibs AMS should store the past 30 days of forecast provided by ibs Analytics in order to give to users an expected accuracy over one month period, well suited for pricing strategy definition purpose.

ENHANCE FORECAST ACCURACY

Based on these accuracy levels, ibs AMS provides the following features:

  • "Passive" use of the expected accuracy:
    • The accuracy levels are exposed to users and retrieved in both graph and grid view. It allows business users to estimate risk levels in defining one or another policy.
    • The accuracy levels are consolidated over KPI dedicated reports for an overall view of the ibs Analytics forecast engine efficiency. Native reports are embedded (by category, by level of sales, by region) and additional ones can be configured.
  • "Active" use of the expected accuracy:
    • Based on accuracy levels provided, identified users can be warned (for example a drop of 5% in accuracy of a given category) in order to focus on it and update forecast models accordingly.
  • "Automated" use of the expected accuracy:
    • the ibs Analytics meta-forecast engine takes full benefits of the estimated accuracy level in order to automatically detect and correct observed bias in forecast for specific profiles of sales.

This advanced and complete tracking of accuracy levels is the key point to provide transparency to the business users and ensure an optimized usage of both forecast and optimization simulations.