Online Behavioural Advertising (OBA) enables promotion companies to effectively target users with ads that best satisfy their purchasing needs. This is highly beneficial for both vendors and publishers who are the owners of the advertising platforms, such as websites and app developers, but at the same time creates a serious privacy threat for users who expose their consumer interests. In this paper, we categorize the available ad-distribution methods and identify their limitations in terms of security, privacy, targeting effectiveness and practicality. We contribute our own system, which utilizes opportunistic networking in order to distribute targeted adverts within a social network. We improve upon previous work by eliminating the need for trust among the users (network nodes) while at the same time achieving low memory and bandwidth overhead, which are inherent problems of many opportunistic networks. Our protocol accomplishes this by identifying similarities between the consumer interests of users and then allows them to share access to the same adverts, which need to be downloaded only once. Although the same ads may be viewed by multiple users, privacy is preserved as the users do not learn each other’s advertising interests. An additional contribution is that malicious users cannot alter the ads in order to spread malicious content, and also, they cannot launch impersonation attacks.