The most used global sensitivity analysis (GSA) method is based on variance. This is performed using Monte Carlo Sampling (MCS) or Latin Hypercube Sampling (LHS). It requires a large sample to obtain accurate estimates. Density-based methods, such as the GSA PAWN, have been developed to reduce the sample size without compromising the result. PAWN is simpler than other methods because it uses cumulative density functions (CDF) instead of probability density. This method has been widely used in areas such as environmental engineering with very good results, reducing computation time. However, its use in structural engineering is incipient. The PAWN method was used to classify the design variables of the isolation system in relation to their sensitivity, and in relation to the seismic response of industrial storage racks. The above was analyzed in terms of the effectiveness of each variable to reduce the seismic demand using a novel base isolation kinematic device (BIKD). Racks with different combinations of their structural parameters such as the number of storage levels, the height between them, and isolation period, among others, were studied. The dimensions of the racks were chosen to match those that would later be experimentally tested on shaking table. An earthquake whose response spectrum matched the design spectrum of current Chilean regulations, was considered as seismic forcing. The maximum base shear load, the displacement of the top level of storage and the floor drift were considered as target responses to be studied. Fixed base racks (FBR), as reference, and base-isolated racks (BIR) were analyzed. The results showed the effectiveness of using the BIKD system in reducing all three-target responses up to one order of magnitude. Additionally, it was determined that the parameters that have the greatest influence on the response correspond to the number of storage levels and the height between them, both for FBR and BIR.