This paper analyses customers’ demand flexibility in a local power trading scenario through an Ising spin-based model. We look at quantitative information on the two-way relationships between power exchanges and spin dynamics. To this end, a modified version of the Metropolis-Hastings algorithm was implemented, including a gradient descent through the constraint surface. This allowed us to analyse the system on a large scale (considering the cumulated benefit of all the actors involved) and also from the perspective of total aggregation. In a maximum flexibility scenario, the total aggregation profit increases with the number of aggregators. We also investigate numerically the effect of aggregator boundaries on the spin dynamics.