Joint publications can noticeably speed up the growth of the h-index, yet in a number of cases they distort its value. We examine how co-authorship is reflected in this scientometric indicator, why the standard formula does not account for each researcher's contribution and which alternative metrics help to gauge an author's real input.