Many of the financial data statistical institutes require from companies are kept in their administrative systems. Statistical agencies, together with other institutes are of course aware of this and so a multitude of formats have been developed to receive the required information in file format, instead of entering the data in a digital form. The format and content of these data need to be highly predictable to enable automated processing. Elaborate message implementation guides and infrastructures serve this purpose. In general, statistical agencies by themselves do not have sufficient leverage towards software vendors to invest in the implementation of such infrastructures. The same goes for the maintenance of the mapping between the administration and the content part of the data file. A standard format and predictable content enable the importing of administrative data in digital forms. XBRL and taxonomies based on reporting standards offer such a standard and the predictability of the content. By its nature, XBRL, does not necessarily result in reports that are complete, i.e. contain all report elements. Although additional rules can be implemented to ensure that complete (XBRL) reports are sent in to regulators, a different approach can also be taken. In this approach reporters (re-)use the XBRL data that is available from administrative systems (or earlier reporting activities) by importing these data into a web form that is made available by the regulator. The web form automatically fills the corresponding fields with the data available in the XBRL document. The fields for which no information is available in the imported data file remain empty. The reporter can subsequently enter the remaining data into these fields manually. Importing data in digital forms would make re-use of data by government regulators transparent. It would visibly reduce administrative burden. Such an approach is taken by the Belgian government. For both the Central Balance sheet office, (annual reporting,) as well as the statistical office, reporting in XBRL is obligatory. Web forms provided by these regulators offer an additional interface between reporter and regulator. In those web forms, the data from the XBRL file are displayed in the corresponding fields. If incomplete or incorrect, the form can be completed manually. At Statistics Netherlands we also experimented with such an approach. The results were encouraging: based on reporting taxonomies, pdf forms could automatically be generated, and XBRL data could be imported into the form. Both approaches are discussed in this presentation. We will discuss how statistics Belgium migrated from digital reporting to XBRL reporting. We will also discuss why their approach could work so well in Belgium, and would not work well in other countries. Alternatives are discussed and a path towards implementation of hybrid web forms is presented.