SUMMARY Food security is threatened by climate change, with heat and drought being the main stresses affecting crop physiology and ecosystem services, such as plant–pollinator interactions. We hypothesize that tracking and ranking pollinators' preferences for flowers under environmental pressure could be used as a marker of plant quality for agricultural breeding to increase crop stress tolerance. Despite increasing relevance of flowers as the most stress sensitive organs, phenotyping platforms aim at identifying traits of resilience by assessing the plant physiological status through remote sensing‐assisted vegetative indexes, but find strong bottlenecks in quantifying flower traits and in accurate genotype‐to‐phenotype prediction. However, as the transport of photoassimilates from leaves (sources) to flowers (sinks) is reduced in low‐resilient plants, flowers are better indicators than leaves of plant well‐being. Indeed, the chemical composition and amount of pollen and nectar that flowers produce, which ultimately serve as food resources for pollinators, change in response to environmental cues. Therefore, pollinators' preferences could be used as a measure of functional source‐to‐sink relationships for breeding decisions. To achieve this challenging goal, we propose to develop a pollinator‐assisted phenotyping and selection platform for automated quantification of Genotype × Environment × Pollinator interactions through an insect geo‐positioning system. Pollinator‐assisted selection can be validated by metabolic, transcriptomic, and ionomic traits, and mapping of candidate genes, linking floral and leaf traits, pollinator preferences, plant resilience, and crop productivity. This radical new approach can change the current paradigm of plant phenotyping and find new paths for crop redomestication and breeding assisted by ecological decisions. Significance Statement While climate change threatens food production through affecting plant physiology and pollinating ecosystem services, plant phenotyping platforms aim at identifying traits of resilience mostly based on vegetative indexes. However, since flowers are better indicators of the plant well‐being and provide food for pollinators, we propose to use the senses of animal pollinators and their preferences to assist in the identification of flower traits of resilience to abiotic stresses for plant breeding based on ecological decisions.