We have a 2yr data and we want to forecast 4 months ahead. But since it's too short, the models doesn't fit really well (20% MAPE), but that's the best we can do so far. But even if MAPE is low, I'm assuming that forecasts are still not perfect and actual values may differ significantly.
Any advice how to present forecasts with a caution to clients that the forecasts may be off?
When actual numbers do come up and forecasts were wrong, what has been your experience on the consequences of that in the industry? McDVOICE
