The prediction of construction productivity is an integral part of the planning and cost estimating process. Although different methods are available and some comparisons between different methods can be found in the literature, an analysis of the data required for productivity estimation with the purpose of developing a unified database of productivity efficiency multipliers is rarely addressed. This paper examines the data required for productivity estimation according to Komatsu and Caterpillar performance manuals with the purpose to analyze dependencies between the data, formulate functional dependencies and investigate the hypothesis that a unified database for the two methods can be designed. For the purposes of this paper only excavation has been analyzed but it is argued that the same methodology can be extended to other construction activities also. The process revealed “dark” areas (i.e. not existing data) in the published data (e.g. for swing angles of less than 45° in the prediction of excavation productivity for Komatsu), different values for the same factor (e.g. different fill factors for the same soil) and, even different reliance between factors (e.g. cycle time for excavation).