Increasingly, microbial production processes are being improved by targeted approaches. In directed strain improvement, the selection of the relevant targets is the limiting step in metabolic engineering. Currently, the identification of leads is still a random process relying largely on expert knowledge. Recently, this approach has been complemented by metabolic flux and control analysis approaches. However, both are closed approaches, and biological processes or interactions that are not currently known to exist, or to be important for bioproduct formation, are not taken into account. By contrast, the recently introduced functional genomics technologies enable an open approach towards target selection. In the near future, we might see that metabolomics, and its integration with transcriptomics and/or proteomics into a systems biology approach, in combination with multivariate data analysis tools, will become of increasing importance for the unbiased selection and ranking of targets, not only for strain improvement but also for bioprocess improvement.