This research presents an analytical model that aims to pin-point influential posts across a social web comprised of a corpus of posts. The model employs the Latent Dirichlet Al-location algorithm to associate posts with topics, and the TF-IDF metric to identify the key posts associated with each top-ic. The model was demonstrated in the domain of customer relationship by enabling careful monitoring of evolving "storms" created by individuals which tend to impact large audiences (either positively or negatively). Future research should be engaged in order to extend the scope of the corpus by including additional relevant publicly available sources.