DMS knowledgebase

1. Forecast at the right level

It is advisable to periodically review your current level of forecast and decide whether it is appropriate given the forecast accuracy goals and review process. There are many options to consider when designing your forecast database. Is a forecast at SKU level appropriate? In many environments it is not and greater detail must be made available by segmenting history and forecast by channel, sales region, or even customer. In other cases companies are currently forecasting at customer level when a more accurate forecast could be achieved with less effort by working at a higher level.  

2. Review forecasts at aggregate levels

When reviewing item level forecasts it is all to easy to “pad” each item’s forecast just in case. It is not until the item level forecast is aggregated and reviewed at a brand or product category level that the cumulative effect of this “padding” is exposed in the form of a clearly unachievable growth in forecast over history. Aggregate level forecast review is an essential part of the forecast review process because it allows for a “sanity check” of the forecast compared to history and preferable company budgets. Any anomalies must be identified and corrected before putting the forecast into the inventory planning system.  

3. Involve the “right” people

The “art” of forecast management involves getting the people who have market information easy access to input their intelligence into the forecast. Their local market or product line knowledge must be tapped into because it will provide valuable information on upcoming demand spikes and troughs. Many of our clients use the Demand Solutions™ Feedback tool to solicit forecast information from sales, product managers, or even customers directly. Whatever mechanism is used to gather the forecast intelligence it must be timely, systematic, and allow for analysis of forecast accuracy.  

4. Review forecast by exception

I always cringe when people tell me they review the forecast by starting at the first product and scrolling to the last. Why not use ABC analysis to focus on the products which are most important first while the mind is still fresh? Why not use deviation filters to identify the few products in the database which have unusually high or low trending forecasts when compared to history? This management by exception will keep you focused and allow for a much more efficient review of the forecast.  

5. Measure and report forecast accuracy

Demand Solutions™ allows you to review and report on forecast accuracy in almost any way you can imagine. Standard reports offer analysis at detailed and summary levels to help you identify weak areas and opportunities for improvement. A forecasting process will rarely be successful if the progress is not measured and the results reported to all stakeholders. Take advantage of the utilities available and start your forecast accuracy graph this month. Let everyone know how “good” or “bad” they are doing and what progress they have made over time. In my experience as soon as a forecast accuracy graph is started the overall trend is up for six months. After that time everyone is committed to the process and the real work can begin to fine tuning the already successful process. For more information on how these ideas can be applied in your environment using the tools in Demand Solutions™ feel free to call our helpdesk and speak with a consultant…and most importantly remember: SAVE YOUR FORECAST EVERY MONTH