Central concerns of research and innovation policies under uncertainty and systemic complexity are market and systemic failures, i.e. both market corrections as well as co-ordination and coherence are essential elements of policy making. Therefore, policy making cannot be based on the notions of equilibrium and optimisation but on adaptive policy making with central focus on search for and identification of policy alternatives and with creation of new policy understanding about functions, underlying mechanisms and conditions of successful policies. For this purpose, collective benchmarking of research and innovation policies provides routines for exploiting various types learning sources”. While benchmarking is a flexible instrument it nevertheless needs a rigorous benchmarking methodology that guides search for best practises and respects the systemic and country specific foundation. Collective benchmarking utilises different types of policy potentials which increase the effectiveness of research and innovation policies and, therefore, the performances of particular countries and the group as a whole.