Energy efficiency programme monitoring is often based on ex-ante saving estimations. These estimations lack standardization, are prone to parameter errors, and neglect rebound effects. Moreover, a comparison with ex-post estimations is not systematically done. This paper proposes a simple methodology to assess ex-ante against ex-post estimations, using the distribution of the energy savings deficit, i.e. the normalized difference between reported and metered savings. This methodology is applied to a dataset of lighting renovation measures to first assess calculated savings against utility meter data. The mean energy savings deficit amounts to 39% for measures in small to medium enterprises and to 28% in common areas of buildings. Further analysis shows that this overestimation arises from both significant calculation errors and rebound effects. These analysis and findings are the first of their kind for these sectors. The methodology is further used to assess deemed savings approaches. The results show that coefficients based on time-of-use and the type of replaced technology provide a precision comparable to calculated savings. Further limitations of ex-ante quantitative monitoring are identified based on empirical observations of a utility-driven programme. This suggests that, instead of current practices, quantitative monitoring should focus on ex-post analysis of meter data. This would free up resources for a more qualitative follow-up of measure impacts to better inform programme and policy design.