One major goal of the European Union’s (EU) Better Regulation agenda has been the introduction of transparent consultation procedures where citizens, interest groups, and businesses alike can voice their opinions about upcoming legislative procedures. Yet, since the Commission ultimately needs to reconcile these diverse preferences and interests in one coherent policy proposal, the question of whose positions prevail in public consultations under which conditions is highly relevant. In the existing literature, lobbying success has been explained by individual-level factors, such as an actor’s resource endowment, or contextual factors, such as the complexity of the policy in question. In this paper, we argue that these two aspects should not be separated and suggest that complexity acts as a moderator and preselection mechanism between individual-level factors and lobbying success. In particular, we expect that, first, complexity raises the information needs of the Commission. Actors that possess the necessary resources can provide this information and are hence able to steer the Commission proposal into their direction in complex settings. Second, complexity raises the costs of participation, incentivizing lobbying actors to self-select into settings where they deem their chances of success higher. We test these propositions on a novel dataset containing more than 700 consultation responses and develop a novel measure for lobbying success based on document embeddings. We show that the measure outperforms prior automated approaches significantly and allows us to study lobbying success in a large-n setting.