Abstract
This paper aims at investigating the relations between supply chain design decisions (i.e. number of supply chain levels, number of nodes at each level, number of sources for each node and distance between nodes) and supply chain performance (i.e. stock-outs at the retailer level) in a pull-based supply chain. A framework expressing the hypothesized relations among the above-mentioned variables has been developed and validated by applying simulation techniques and statistical analysis. The number of nodes at each level has been demonstrated to increase stock-outs at the retailer level. On the contrary, performance is not affected by both the number of supply chain levels and the distance between nodes. The number of sources for each node seems to increase stock-outs at the retailer level but no statistical evidence for this has been found. This work can support managers in taking supply chain design decisions and in defining countermeasures to mitigate their effects on supply chain performance.