The control of antibody specificity plays pivotal roles in key technological fields such as diagnostics and therapeutics. During the development of immunoassays (IAs) for the biosensing of pathogens in food matrices, we have found a way to rationalize and control the specificity of polyclonal antibodies (sera) for a complex analytical target (the Salmonella genus), in terms of number of analytes (Salmonella species) and potential cross-reactivity with similar analytes (other bacteria strains). Indeed, the biosensing of Salmonella required the development of sera and serum mixtures displaying homogeneous specificity for a large set of strains showing broad biochemical variety (54 Salmonella serovars tested in this study), which partially overlaps with the molecular features of other class of bacteria (like specific serogroups of E. coli). To achieve a trade-off between specificity harmonisation and maximization, we have developed a strategy based on the conversion of the specificity profiles of individual sera in to numerical descriptors, which allow predicting the capacity of serum mixtures to detect multiple bacteria strains. This approach does not imply laborious purification steps and results advantageous for process scaling-up, and may help in the customization of the specificity profiles of antibodies needed for diagnostic and therapeutic applications such as multi-analyte detection and recombinant antibody engineering, respectively.