On many roads in rural and mountainous areas, the cellular network connectivity is intermittent and dead spots, i.e., zones without any coverage, are frequent. In previous work, we developed a data dissemination protocol to accelerate the transmission of messages in dead spots. It combines the cellular network with short-living ad-hoc networks between vehicles. A car in a dead spot can forward messages directed towards the environment, to the peer in its ad-hoc network that will leave the dead spot first, effectively reducing the delay. An issue, however, is to reliably identify the peer that is most likely the first one regaining cellular network coverage. This problem can be solved if the borders of the dead spot, the vehicles are in, are previously known. For that, we use a novel technology named dead spot prediction. Here, vehicles conduct local connectivity measurements that are aggregated to so-called connectivity maps describing the locations of dead spots on a road system. In this article, we introduce the combination of the data dissemination protocol with dead spot prediction. Particularly, our protocol is amended such that connectivity maps are considered when deciding which vehicle leaves a dead spot first. Since currently only few publicly available works about dead spot prediction exist, we further created a prototype of such a predictor ourselves that will be discussed as well.