In the study, we focus on examining the impact of selected factors on the results of electronic reverse auctions and thus the savings achieved. The study extends the knowledge of extant literature to empirical insights and uses currently the broadest sample of approximately 16,000 e-auction cases. The study confirms that the savings calculation method and the use of a comparative or initial price affect the outcome of the auction as well as the individual selected factors entering the model (e-auction value, number of changes, number of extensions, number of bidders, number of items).