Foragers within a group might increase individual foraging efficiencies by using public information to assess local resource availability. This information is often expressed as a change in behavior at resource encounters, which can be detected by nearby individuals. When the resource landscape displays sufficient degrees of clustering or fractality, joining nearby conspecifics in their successful foraging efforts becomes an attractive strategy. However, when group sizes increase, joining others might become less effective due to increased levels of intraspecific competition. In this work, we introduce a trimodal Lévy search in fractal resource landscapes, where highly diffusive, extensive searches are interchanged with localized, intensive searches and informed searches guided by attraction to successful conspecifics. Using an agent-based model, we are able to quantitatively determine what environmental characteristics facilitate joining to be beneficial on both the group and the individual level. We find that joining others is advantageous on the group level, but only if resources are sufficiently clustered and group sizes and joining ranges are not too large. In contrast, individual advantages, expressed as an increased survival rate mediated by a reduced variation in resource intake rate, are largest precisely in parameter regions where group benefits were smallest. These results highlight both the reach of a relatively simple agent-based model based on multimodal random searches, and the notion that groups might not necessarily optimize foraging efficiency when more strict conditions, such as survival, need to be met.