Its better to unpack what Tableau is telling you - your data is too volatile in order to calculate a reliable forecast.
Looking closely at the forecast, this looks like it was caused by the rise from 0 to 200k in the 2013-14 period, and its remained within the 200k-400k band until now. Using this logic, you could expect that a forecast EXCLUDING the 2013-14 period should remain within the 200k-400k spectrum.
If we make an assumption on the model, that 2013-14 was a setup/growth phase and not truly representative of normal business, then you could fairly safely remove it from your forecast model. In Tableau, you can do that easily by just filtering out the dates.
Thanks Pete. I understood now I know what to do. I really appreciate your advise.
Thanks Shini for helping me to optimize the model.