2 of 2 people found this helpful
It looks like you're getting a flat line forecast for some categories and just trend for others as we're not able to find any seasonal cycles. Does your data have a strong seasonal component? Also could there be another variable that is affecting this variable. Tableau's built-in forecasting is univariate, meaning it looks at previous values of X to predict future values of X. But there could be other variables A,B,C that impact X. E.g. sales can be impacted by consumer confidence, unemployment, population... Or you can simply have events that cause abrupt changes e.g. black friday, christmas... You can do this kind of forecasting also in Tableau but you'll need to rely on R integration.
I have put together an example workbook here in case you'd like to take a look
Thank you for the detailed answer. I have used R a lot in the past and am excited to learn about the R integration.
What I've noticed is that if I select the view at Quarters, it gives reasonable estimates that are not flat compared to monthly analysis. I think it's possible that the forecasting model is just not good at predicting intervals as small as months and is better suited for quarters which has a better fit with a seasonal models.