A researcher's number of publications has been a fundamental merit in the competition for academic positions since the late 18th century. Today, the simple counting of publications has been supplemented with a whole range of bibliometric indicators, which supposedly not only measures the volume of research but also its impact. In this study, we investigate how bibliometrics are used for evaluating the impact and quality of publications in two specific settings: biomedicine and economics. Our study exposes the various metrics used in external evaluations of candidates for academic positions at Swedish universities. Moreover, we show how different bibliometric indicators, both explicitly and implicitly, are employed to assess and rank candidates. Our findings contribute to a further understanding of bibliometric indicators as "judgment devices" that are employed in evaluating individuals and their published works within specific fields. We also show how "expertise" in using bibliometrics for evaluative purposes is negotiated at the interface between domain knowledge and skills in using indicators. In line with these results, we propose that the use of metrics we report is best described as a form of "citizen bibliometrics"--an underspecified term which we build upon in the article.