This talk deals with strategic voting under incomplete information where agents can iteratively change their voting intention, based on the information they receive from their environment. We propose a descriptive model, inspired by political elections, where the information that the agents receive comes from public opinion polls and social networks. The voters are assumed to be confident in the poll and they update the results communicated by the polling agency with the information they get from their relatives in the social network. In this context, we investigate the manipulative power of the polling agency, by analyzing both constructive and destructive manipulation. The theoretical results are complemented by experiments using different heuristics.