https://demand-planning.com/2012/05/31/the-science-of-forecasting/ S&OP/ IBP, Demand Planning, Supply Chain Planning, Business Forecasting Blog Thu, 31 May 2012 18:46:57 +0000 hourly 1 https://wordpress.org/?v=6.6.4 https://demand-planning.com/2012/05/31/the-science-of-forecasting/#comment-214 Thu, 31 May 2012 18:46:57 +0000 https://demand-planning.com/?p=1369#comment-214 Yes indeed I would agree to begin with Random Walk, then consider a Seasonal Random Walk, then maybe a Trended Seasonal Random Walk, and, with a front loading. (Forecast is about date and quantity, not just a quantity per period)
But first, be sure you clean the data. A 95% service level probably means the historical data has 5% holes in it, caused by Out of Stock, or Credit limit situations. Or, it has bumps in it caused by Blow Outs, Promotions, and recovery from Out of Stocks, and etc.
Then consider process and procedure, of how this forecast is to be applied, and maintained.
The FVA approach can help with all of the above.

]]>