Lately, we have seen that the behavior of the different type of supply chains in different industries varies a lot in case of managing demand. Make to stock style companies tend to focus on demand management using a combination of statistical techniques and collaboration to gain market force knowledge to deliver a forecast for the purpose of setting inventory/ service policies whereas make to order businesses forecast only as a starting point, but more frequently use sales driven “consensus” forecasting with limited statistics, and use the forecast to drive procurement decisions and capacity management, rather than build plans for end items. Moreover, best in class organizations use supply chain techniques to gain control in the demand execution window.
The focus areas for all of them are usually the same:
- Improving demand forecast accuracy using internal demand planning
- Creating customer driven forecasts through robust sales and operations processes
- Creating sell side collaborative processes to manage demand within the demand execution window
These focus areas are important because of the current economic conditions that put pressures like increased volatility of demand and escalating customer service demands. These pressures drive the organizations to keep on improving demand forecast accuracy and developing a more responsive inventory placement strategy to stay ahead of time and to remain best in class. So, how do we improve the forecast accuracy, you might ask this question?
Let us discuss on few steps that I think have the capability to improve the forecast accuracy and brings it closest to the actual demand:
- First of all, we need to utilize the multiple roles within the organization to generate their independent forecasts giving the full set of assumptions that they took
- Include product promotion, competitive pricing, demand shaping events and historical ups and downs into the forecasts
- Collaborate with all internal stakeholders by consolidating the forecasts and having an open forum discussion on what was missed and what can have more impact on the demand in coming time – please note that this should not take more than 15 minutes per product and everybody should come prepared to quickly nail down the numbers
- Collaborate with external stakeholders like customers and/or suppliers
- Integrate trading outbound supply chain partner data with internal process organization
- Ensure that the application that sores the historical data of demand has excellent data accuracy as lot will depend on this data
All above will work for sure in case your organization follows the make to stock strategy but for the make to order organizations, more focus will be on external collaboration like with customers and suppliers so that desired accuracy in the forecasts can be achieved. We have seen different opinions on whether we should follow the make to stock option or got for make to stock option. The summary of the push-pull model of demand management is that the organization finds and continually operates at the optimal operating point between a pure build to stock environment and a pure build to order environment. This optimal operating point varies by model, by market, and over time. Pure build to stock is undesirable because of the cost of deployed inventory and the inevitable mismatches between inventory and what customers want. Even pure build to order is undesirable because of the cost of the deployed capacity required to support highly uneven demand. The optimal solution is somewhere in-between and a blend of both will be the best option to go for.