This paper is on the development of the KDK Thesaurus, a CV partly based on ELSST, used for the topical discovery of interview materials. The presentation discusses the workload for such a project, its sustainability and future perspectives of similar projects, incl. ONTOLISST and touch on the issue of economical and ethical considerations of metadata curation on smaller levels of datasets like parts of interviews and variables. The paper also examines how social science data archives may operate in economically scarce environments and how big of a burden the costs of the introduction of elaborate metadata standards like DDI or the use of different controlled vocabularies might entail.