I have a client which started on december 2014 and they are only capable of sending their sales during middle and last day of the month. I used exponential smoothing to get their forecast sales. The ff is the sample forecast of one of their items with the weighted value of 0.6.

Cutoff Date | Actual Value | Forecast Value
12-15-2014  |       1      |    
12-31-2014  |       61     |       1
01-15-2015  |       24     |       43
01-31-2015  |       5      |       29.7

The forecast value has a big difference to actual value. I read about the seasonal adjustment data from this link https://analysights.wordpress.com/2010/09/16/forecast-friday-topic-forecasting-with-seasonally-adjusted-data/ but it still requires at least a year of data. Is there a way to make the forecast more accurate?

  • $\begingroup$ You could use analogue data/product to capture seasonality. Is this retail industry? If so it is fairly straight forward to get analogue product $\endgroup$ – forecaster Feb 17 '15 at 14:36
  • $\begingroup$ yes it is retail industry, I understand but how does it affect the computation of the forecast? Should a use multiplier base on the data analogue? $\endgroup$ – reggie Feb 23 '15 at 1:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.