This work presents a feature-oriented taxonomy for commercial electronic voting machines, which focuses on usability aspects. We analyze the incremental differences in functionality - the features - between voting systems and how that functionality differs from ‘traditional' pen and paper based voting. Based on this analysis, we propose a ‘Just-Like-Paper' (JLP) classifcation method. We validated the JLP classifcation method to the ongoing development of our novel DualVote eVoting system, where its application led to the development of a passive feedback protocol which improved the usability features