Handling Forecast Outliers

Greetings Anaplan Community, 

We are currently using the Anaplan statistical forecasting model for our demand planning, and a question has come up around how we can better handle outliers that can drastically skew our forecasts when there are periods of zero demand. We have a process for manually backfilling demand in these cases, but the business is looking for a control on the forecast reactivity that would handle these in a more automated fashion. Has anyone encountered a similar problem, and if so, what methods have you put in place to handle outliers?

Thank you!



  • ankit_cheeni
    edited February 16

    Not sure how helpful this would be but we faced a similar issue for a Liquor Manufacturer. We stuck to the basic and did a 'Mean ± 2 Standard Deviation', mean and Standard deviation calculated within the system. Worked fine for us.