We certainly agree with Schradin [1] that comparative studies need to use the best information available—as well as clear definitions, up-to-date statistical techniques and common sense. We have divided our response into two sections: the first dealing with Schradin's critique of our 2017 paper and the classification of the social organization of shrews; the second dealing with three general issues that his commentary raises. The paper that Schradin is commenting on [2] asks whether cooperative breeders (which we define as species where non-breeding helpers assist breeders to raise their young) in mammals are more commonly found in arid environments than monogamous species (defined as species where breeding pairs of males and females remain together for more than one season) from which cooperative breeders appear to be derived [3]. It shows that, as in birds, cooperative breeders tend to live in relatively arid habitats compared to monogamous species. As we describe below, these conclusions are unaffected by differences between our classification of the social organization of shrews and the classification suggested by Schradin. Schradin's first criticism is that the categorizations of shrews that we have used in this analysis were based on one paper on a single species. This is incorrect, for our categorization of shrews was based on more than 70 separate sources (see electronic supplementary material, appendix S1). When we published our dataset [4], we were required to reduce the associated reference list by the editors of the journal and, to comply, we removed around 75% of our 2000+ sources, often listing a single reference per species or group of species, and frequently using generic reviews for particular groups where these were available: for example, for many rodent species, we simply listed Wolff and Sherman's Rodent Societies [5], and …