High-performance and detailed insights into model and forecast quality

The ability to accurately forecast future sales impacts all areas of an organisation - from strategic planning, purchasing and supply chain management to individual shops. Retailers face the challenge of accurately predicting which products will be needed in which shop at which time to ensure seamless availability of the desired products for their customers. Although theoretically all the data is available for an accurate forecast, the quality of the forecasts often falls short of expectations. This is due to unrecognised errors in the forecasting model, which affect the accuracy of the forecasts.


This is why retailsolutions has developed the SAP Analytics Suite add-on. In addition to SAP Unified Demand Forecast (UDF), this content suite developed specifically for the retail sector offers interactive dashboards that show the problem areas at a glance.  Based on SAP HANA and SAP Analytics Cloud, the suite visualises critical article-store combinations using various error KPIs. This highly aggregated view of the underlying issues enables users to take appropriate action to improve forecast quality and thus effectively respond to the changing needs of their customers.

rs_Product

  • Development of the content suite in SAP HANA and SAP Analytics Cloud (SAC)

  • Use of UDF data to analyse forecast and model accuracy

  • Interactive dashboards to display critical article-branch combinations based on various error KPIs

  • Clear visualisation of the relevant UDF data

  • Simulation of multiple configuration data to compare parameter changes

Zahnraeder_2

  • Highlighting of problem areas and breakdown to detailed level

  • Quick response and problem solving 

  • Connection to the ‘Analyse Forecast’ app for additional information

rs_Benefits

  • Improved forecasting accuracy by analysing various KPIs

  • Ability to quickly identify critical article-store combinations

  • Efficient comparison options thanks to the simulation of configuration data

Forecast_Monitor

Forecast Monitor

Monitoring the accuracy of specific products, locations or hierarchy elements. Supports the validation of forecast quality prior to implementation and ongoing maintenance.

Artikelhierarchie-Analyse

Product hierarchy analysis

Analysis of the product hierarchy to direct planning to the products and locations with the lowest forecasting accuracy.

Analyse_der_Lokationshierarchie

Analysis of the location hierarchy

Analysis of the location hierarchy to compare the genuineness of nodes with each other, with drilldowns in the product hierarchy.

Ereignis_Analyse

Analysis of incidents

Analysis of model accuracy, especially across events, to analyse the accuracy of the model. Comparing multiple diagnostic IDs to validate the performance of UDF parameter settings.

Batch_Monitor

Batch Monitor

Monitoring of UDF modelling, forecasting and HPR batch jobs and analysis of runtime, memory and CPU trends in comparison to each other. Enables proactive monitoring of system performance to detect potential problems, e.g. with growing data volumes, at an early stage.