In breeding programs, balancing short-term genetic gain and loss of diversity per generation is essential to sustain a long-term genetic response. Depending on the dynamic of the species, the acceptable trade-off will be different. One of the most common and successful tools to achieve this management is the Optimal Contribution Selection (OCS), which readily mathematically formulate the trade-off between genetic gain and coancestry. However, OCS only accounts for the next generation gain and diversity, which can lead to suboptimality given the uncertainties of random mating and segregation. In this paper, we have extended the OCS by conveniently integrating a way to promote certain parental pairs, so that this method can account for the next t+2 generation. In the study case of Populus nigra, fully phenotyped and SNP array genotyped, we have shown that (i) a non negligible part of the long-term success of a breeding strategy depends on the implemented mating strategy, and (ii) favoring a compensatory mating can accelerate the selection without compromising the future diversity.