In Zendesk forecasting, smoothing parameters are crucial for processing forecasts. These parameters include alpha (α) for the level, beta (β) for the trend, and gamma (γ) for seasonality. They determine the importance of recent versus past observations in the time series.
For example, a low alpha value indicates that the level estimate is based on both recent and distant past observations. A beta value of 0.00 means the trend component's slope remains constant over time. A high gamma value suggests that the seasonal component is based on very recent observations. Zendesk Explore uses its own algorithm to optimize these parameters for accurate forecasting.
To add a forecast to your Zendesk report, you need to create a time-based report first. In the report builder, add metrics like COUNT(Tickets) to the Metrics panel and a date-based attribute like Ticket created - Year to the Columns panel. Then,…
Zendesk Explore offers two forecasting methods: the AA (Additive trend and Additive season) model and the AM (Additive trend and Multiplicative season) model. The AA model is the default and is commonly used for producing realistic results. It…
If you encounter the 'forecasterroryear_missing' error, it means your report is missing a year-based attribute. This error occurs when forecasting is based on recurring attributes like day or month, causing data aggregation issues. To resolve…
The Holt winters model used in Zendesk forecasting is a triple exponential smoothing method that considers the level, trend, and season of a time series. It helps in making accurate predictions by accounting for seasonal variations. Zendesk Explore…
The 'You need to have data for at least 2 cycles' error occurs when your data does not cover enough cycles for forecasting. To resolve this, ensure that your report includes at least two complete cycles of data. For example, if you are forecasting…