Food webs are a unifying concept that spans population, community, and ecosystem ecology. Yet there is a fundamental dissonance between dynamic food webs, which derive from community ecology and characterize the effect of one species upon another, and energetic-based food webs, which have their roots in ecosystem ecology and characterize the flow of energy and matter among species and trophic levels. Here, we present a framework that explicitly defines food web linkage strength in terms of two key factors: the type of data (energetic vs dynamic), and the trophic perspective (whether the interaction is viewed from the perspective of the consumer or the resource). As a case study, we applied this framework to the well-studied trophic interaction between rainbow smelt Osmerus mordax and cisco Coregonus artedii in Sparkling Lake, Wisconsin, USA. Energetic and dynamic metrics gave different descriptions of the linkage strength from the perspective of either species. This was particularly true from the cisco perspective: the dynamic metric indicated a strong linkage because smelt extirpated cisco, but the energetic metric indicated a weak linkage because smelt prey on cisco larvae and the flux from cisco to smelt was therefore a small fraction of cisco production. Our approach highlights the need for more careful consideration of how food web linkages and their importance are characterized. Furthermore, it can provide the basis for translating energetic data to an understanding of food web dynamics.