Sales Intelligence – Simulate Recommendations

If setting “Sales Intelligence – Simulate Recommendations” is set to Yes, several new fields are displayed, that are used to setup the simulation of the recommendation analysis. This feature is used to test the recommendation model on real business data to experiment and assess the accuracy of different parameter setup. This helps to find the parameter values that gives the best item recommendations for a specific business. Default option is No.

Many of the settings are the same as can be found in section Sales Intelligence – Settings. Click COPY CURRENT SETTINGS to copy the field values to section Sales Intelligence – Simulate Recommendations. Please note that parameter “Days back to find relations” is calculated backwards from the date set in “To date” for the simulation, and not from current date.

Two additional parameters are available to set up the simulation:

• To date: This setting is used to define the time interval for the data analysis. The starting date is defined with settings “From Date” or “Days Back to find relations”, and this setting defines the end date.

• Number of days of simulation: This setting is used to setup the number of days to use when running the simulation on actual business data. This is counted from the date setup in setting “To date”.

Click RUN SIMULATION to run the simulation using the parameter setup in section Sales Intelligence – Simulate Recommendations.

The result of the simulation is presented as a Hit Rate % for three different setups: without trends, using trends with separate periods and using trends with accumulated periods. This enables the user to choose the most accurate method of analysis given the values for the other input parameters. Please note that if trends are disabled in section Sales Intelligence – Simulate Recommendations, “Number of days of simulation” is set to 30 days to run the simulation with trends.