Demand-driven Forecasting

 

   Forecasting is the process of making statements about the outcome of future events. The Oxford dictionary defined forecasting as a process of predicting future events based on past or current information. Experts or planners of an organization make decisions based on some forecasts such as predictions on seasonal demands or sudden changes in demand levels after a pandemic. In a competitive market, forecasting can help organizations gain some competitive advantage through innovation. However, it is critical to choose a forecasting technique for your needs. 

   There are different forecasting techniques available such as time series analysis, regression analysis, stochastic methods, relevance trees, and morphology (Daim et al, 2006). An accurate forecast helps to make informed decisions and increase sales. Nestle, a food manufacturing company, saved millions of dollars from their inventory by eliminating the human judgment factor and implementing demand-driven forecasting solutions. Charles Chase, an expert in sales forecasting, was the hero that refined their sales strategies to increase revenue. According to Chase (2013), a company must be proactive versus reactive; experts should not react to the forecast but proactively drive the forecast through demand sensing and demand shaping for thousands of similar products. This iterative exercise should be completed automatically or programmatically for faster results. Chase went further and stated that experts should combine data analytics and domain knowledge and practice lean forecasting and forecast value-added seasonally. In a demand-driven forecasting model, it is important to sense demand signals associated with different sales strategies such as promotions and advertising to better understand what drives consumers towards your products. Once in possession of the data, experts can now start reflecting on future opportunities by constantly asking themselves “What if…” daily.

 Figure 1 shows the differences between traditional and demand-driven supply chains. The demand-driven model is not only driven by customer demand but also the pull technique to get real-time information and more realistic provisioning. 

Figure 1: Traditional chain vs Demand-driven supply chain

   Although a demand-driven forecasting model can be successful in predicting future sales, it is important to recognize some of its limitations such as skewed judgment. As stated by Chase, about 80% of forecasts by Nestle were influenced by human judgment on every cycle. Only 20% of the process was mathematically driven which is considered a red flag. Humans tend to adjust data to align with what they want and not the reality. According to Chase, human interference should be removed from the process. Does this demand-driven model work for any type of product?

References

1. Gosasang, V., Chandraprakaikul, W., & Kiattisin, S. (2011). A comparison of traditional and neural networks forecasting techniques for container throughput at Bangkok port. The Asian Journal of Shipping and Logistics, 27(3), 463-482.
2. Daim, T. U., Rueda, G., Martin, H., & Gerdsri, P. (2006). Forecasting emerging technologies: Use of bibliometrics and patent analysis. Technological Forecasting and Social Change, 73(8), 981-1012.
3. How demand-driven forecasting paid off for Nestle: Logistics. Supply Chain Digital. (n.d.). Retrieved October 17, 2021, from https://supplychaindigital.com/logistics-1/how-demand-driven-forecasting-paid-nestle.