The Slovenian Social Science Data Archives (ADP) has identified the Dataverse application as the best new option for distributing its curated catalogue, which is currently hosted on its website and on Nesstar. In addition, the application was also seen as the best way to provide additional services for self-archiving and basically as an ingest point for all studies added to the catalogue (e.g. curated and self-archived). However, it soon became apparent that the default version of Dataverse lacked many features that were essential to our workflow and the services we wanted. We have invested many resources over the last few years in customizing the default version and have tested several plugins that could be crucial (e.g. Data Curation plugin, Data Explorer plugin, Two Ravens). These include a multi-language user interface, implementation of the CMM CESSDA Metadata Model (and some ADP custom metadata fields), including DDI Controlled Vocabularies and ELSST keywords, and customization of deposit and access features (e.g., inclusion of online deposit agreements and terms of use). The paper will present our experience in customising the Dataverse software application to our needs, and identify issues and workarounds that repositories should be aware of when customising the application to their workflows.