Its fascinating that everyone is a forecasting expert, give them a series of numbers and they will have a go at an estimate of the next 3 in the series. If you ask them to predict the probability of something such as tossing a head or a tail, after an observed series again they will have a go with a modicum of confidence.
There are plenty of algorithms that can analyse historical data and generate a sales forecast, many of them have a pick best logic, to retrospectively generate a best fit view, where as this can generate some reasonable numbers it often generates unstable changes within the forecast.
If your business needs both forecast stability and forecast accuracy, whilst generating an analysis on of demand variability then you probably need best in class toolsthat can look at demand signals at the item by location level, with an analysis at the order/customer level then these tools will give you a forecast with a recommended safety stock to ensure you have an ability to maximise your customer service with your best level of inventory.
If this kind of strategy is something that your business is looking for, your customers are demanding better customer service, your product portfolio is proliferating and your long tail of products has grown significantly, then using best in class tools (systems of innovation according to Gartner) then you ought to be considering companies like ToolsGroup.
What is your strategy for adding capability to your system of record, your existing ERP system?
Would value your thoughts.
Dave Food