The growing importance of the data in today’s applications, such as machine learning, makes merchandising them more appealing, but it opens various challenges. Indeed, we want data to be delivered in a fast, reliable, and secure manner. This means we want guarantees about the payment and the correct delivery of the data, while still offering the ease of existing publish/subscribe protocols. There exists no state-of-the-art protocol that answers to all those challenges. Most of them lack the security properties needed for merchandising data and the few secured propositions do not scale well with the number of buyers, which prevents them from global use. In this paper, we present a data payment system based on SUPRA, a publish/subscribe protocol having delivery guarantees. With our solution, it is possible to securely sell data while sharing them using the publish/subscribe model, which is known for its scalability in one-to-many communications. After presenting how our system works, we explain how it answers the various challenges of merchandising data and compare it to the other state-of-the-art solutions.