This paper is aimed at formalizing and testing a model for explaining the relations among supply chain design decisions (i.e. sourcing strategies, number of network stages, nodes’ capacity and localization) and the need for collaboration between the nodes of pull-based supply chain. As a consequence, a causal diagram which depicts such relations is built and its validation is performed by means of simulation techniques and statistical analyses (ANOVA and linear regression). In particular, since the magnitude of the relations is out of the scope of the paper, the validation is done in relative terms. The results obtained by running the simulation model of the supply chain which represents the ‘base case’ are compared with the outputs of the models which simulates the original logistic network modified according to the above mentioned supply chain design elements. The statistical analyses allow the majority of the relations of the proposed causal diagram to be actually validated.