Abstract Species importance is often defined by how central the role a species plays in a food web. Here we argue that species importance should also incorporate the concept of uniqueness of its network position. By developing a previous methodology, we propose a new approach for species uniqueness. Our methodology quantifies the interaction structure of species, separates strong and weak interactors in a threshold-dependent way and measures the similarity of strong interactors’ identity for pairs of species. By exploring various threshold values systematically, the profile of a species’ trophic overlap can be constructed. Summing up the extent of trophic overlap across the whole profile for a species, we then obtain a new measure for species uniqueness. A unique species should overlap less with other species and therefore has a low value for this new measure. We demonstrate our methodology by using the food web representation of Prince Williams Sound ecosystem. Drastic differences between our new result and that derived from the previous approach are found. We further compare the result with other network indices and found that the information generated from our new approach differs from the majority others. This in turn implies our approach offer an alternative view on species importance that may complement the existing methodologies in the literature.