My manufacturing company has 11 final products. All products have four categories based on sizes. The entire system cannot see the real end-demand. Half of the company’s customers are distributors and not real customers. So the company has to depend on history based forecasts. This results in obvious sales-forecast mismatch. In that scenario how to minimize stock-outs and maximize customer service levels with least network inventory?
Improving operational resilience in a complex production distribution system facing uncertain demand is very challenging task. Firm's operations must be driven closely by customer requests rather than by plans to make and sell. It can be argued that when unpredictability is a given, in the form of mismatch between actual demand and forecast, the only strategy that makes sense is a strategy to become adaptive — to sense early and respond quickly to abrupt changes in individual customer needs.
In order to improve the operational resilience, you must have an adaptive ability to change decisions quickly with respect to changes in environment. Initially, you have to start working in following dimensions:
knowing earlier (demand)
an integrated network ( efficient and well-coordinated system)
small production schedules (small batches)