Evaluating the credibility of information is a difficult yet essential task in a society where the Internet plays a large role. Familiarity with the topic at hand has been shown to have a large influence on the way credibility is evaluated; 'familiar users' tend to focus on semantic features, 'unfamiliar users' focus more on surface features. In this study, we attempt to find out whether these differences have consequences for the development of credibility evaluation support systems. Two simulated support systems were evaluated, one utilizing semantic features (aiming at familiar users), the other utilizing surface features (aiming at unfamiliar users). The results suggest that unfamiliar users have a preference for a surface support system. Familiar users have no clear preference, have less trust in the support, and report to be influenced less by a support system. We recommend to focus on unfamiliar users when developing credibility evaluation support. Copyright 2012 by Human Factors and Ergonomics Society, Inc. All rights reserved.