The performance evaluations of universities and academics are undergoing unprecedented expansion. Excellence in academia might not naturally translate into societal impact; alternative assessment instruments have been developed in recent years, aimed at extending conceptions of academic excellence. New metric techniques like altmetrics play a key role in capturing non-academic engagement of research in a broad range of sources, meant as a complement to traditional indicators. Just as citations are seen as indicative of a degree of academic recognition, references to scientific publications in non-scholarly sources are seen as indicative of a certain societal recognition they enjoy. However, their partial reliance on social media like Twitter makes them prone to manipulations like academic spamming. References to scientific publications cited in policy documents seem a promising avenue for application in research evaluation, as these are presumably more likely to signal a direct SIUR and seem less prone to questionable practices. Novel databases like Overton allow for detailed analysis of research uptake in more than six million policy-related documents from countries worldwide. But because the ability to measure and analyze citations to academic works in policy documents has only recently become a technical possibility, questions remain as to which and whose research appears in those documents. More research is needed that focuses on unexpected biases these objects might generate; moreover, gender and seniority of researchers appearing in Overton have not been thoroughly investigated.