Getting to know new people via social media is a challenge. Users struggle with assessing the trustworthiness of others via the software application. However, believing that other users are trustworthy and act towards one’s positive expectations is oftentimes a prerequisite for interactions - whether to start or continue online interactions or transfer them to the offline sphere. This work investigates the research objective of how software engineers can develop social media applications that support users in their online trustworthiness assessment. This dissertation is based on previous trust research and a conducted literature review to provide a theoretical model for trustworthiness in social media. Furthermore, existing practices in requirements engineering, risk management, goal modelling, and feature modelling are considered, to introduce various practical methods for software engineers. In addition, quantitative and qualitative research is conducted to apply, test, and evaluate the methods of this work. The theoretical results of this work are the trustworthiness framework for computer-mediated introductions and the overview of trustworthiness facets. Both provide a theoretical framework for the practical implications by which psychological processes can be transferred to the online environment. Practical results are the method for eliciting trust-related software features, called TrustSoFt, the guideline for selecting trustworthiness facets, i* goal modelling for TrustSoFt, and the method for establishing feature models for trustworthiness assessment. All practical approaches support software engineers in specifying trust-related software features by which users can assess the trustworthiness of other users, the service provider, and the social media application. The qualitative and quantitative evaluation of TrustSoFt has shown that TrustSoFt is appreciated by users of the method. The resulting software features of TrustSoFt reduce the trust concerns of end users. The TrustSoFt use case for online dating applications resulted in innovative trust-related software features that are not used by existing applications, yet. Future work can facilitate the TrustSoFt application by developing a digital tool. Furthermore, additional qualitative and quantitative studies or practical applications in industrial software development projects can support the validity of the theoretical and practical findings.