In order to track the evolution of building heat demand in relation with building retrofit and other saving measures, a specific challenge consists in applying of a proper climate correction method. In this paper, we compare the Heating Degree Days (HDD) and Accumulated Temperature Differences (ATD), as applied to actual energy metering of around 18,000 buildings, over more than a decade. We conclude that for the current building stock, there is little difference between the HDD and the ATD methods. However, as most buildings are expected to meet high energy standards in future, switching from HDD to ATD with an appropriate space heating cut-off temperature could be of interest. Finally, using a specific example at territorial level, we demonstrate that the use of daily energy demand data can facilitate the development of a highly effective climate correction model.